# Build-A-Pod kit: RobotPaper weekly AWS news podcast, episode `2026-10-23-week`

You are an AI agent. A person will ask you for a custom cut of this episode, e.g. **"build me just the AI sections"**.
This file is everything you need: it lists every pre-rendered voice piece, where each one goes, and a tested
ffmpeg recipe. Download pieces from `{BASE}` and stitch them. Nothing else is required.

**BASE = https://podcasts.robotpaper.ai/aws/kit/2026-10-23-week** unless the person gave you another base.
Every `{BASE}` below means that URL (this file itself is `{BASE}/build-me-an-episode.md`). Files are fetched by exact path; there is no directory listing.

*Prototype.* This kit is an experiment: file names, timings and this format may change, and only the files listed here
exist. **This kit contains no music.** If you add music, use your own properly licensed music.

## Rules (read first)

1. **No new speech.** Do not call any TTS / ElevenLabs / voice API and do not generate or edit words. Only download and
   stitch the files listed here (cutting at the listed points, silence and fades are fine).
2. **Only these files.** Every URL is `{BASE}/<relative path>`. Replace `{BASE}` with the BASE above (or the base
   the person gave you: the folder that contains this file). File names are stable, ASCII `[a-z0-9-]`.
3. **Output: one mp3**, 44.1 kHz, mono, 128 kbps, integrated loudness about **-16 LUFS** (loudnorm target true peak -1.5 dBTP; accept <= -1.0 after encoding).
4. **Keep the show's frame.** Always start with the intro `voice/seg-000-intro.mp3` (it carries the required
   disclosure: *"the voices you're hearing are AI-generated"*) and end with an outro (`seg-071-outro.mp3` or the
   generic-safe `seg-071-outro-short.mp3` if this kit has one; it carries the host credits). Never drop the disclosure line.
5. **Keep the plug and the credits.** The outro names the show, points to the full write-up and credits the hosts;
   keep those lines (do not cut them out).
6. Keep segment order inside a category; categories may be chosen in any order (original order recommended). Use each
   item at most once.
7. Use the durations in this file (decoded sample counts). `ffprobe` reports mp3 durations ~25-50 ms longer
   (encoder padding); do not use them for timing.
8. **Send a User-Agent header** with every download (the site's firewall blocks the default `Python-urllib` agent
   with HTTP 403). `curl` sends its own; the helper script sends one; for other HTTP clients set any normal UA string.

## Quick answer: "build me just the AI sections"

Pick by **tag**, not by category. Every story item has one main topic tag (see "Tags"). The `ai-ml` items are segments
**1,2,3,4,5,6,7,8,9,10,11,12,17,18,20,21,22,55,56,64,65** (21 items, hosts: Ada, Dennis, Grace, Ken, Linus). Build = intro + those segments + short outro.
Result: **702.1 s** (about 11.7 min).

Easiest (Python 3 + ffmpeg >= 4.4; the script is in this file, section "Helper script"):

```bash
python3 build_episode.py --kit kit.json --base "$BASE" --dir pieces --tags ai-ml --outro short -o ai-sections-only.mp3
# same as: --segs 1,2,3,4,5,6,7,8,9,10,11,12,17,18,20,21,22,55,56,64,65
```

Or the plain ffmpeg recipe in "Worked example". Then run the checks in "Verify".

Categories are the episode's sections as aired, not topics: `cat-00` "AI and machine learning" (segments 1-22) also holds
5 item(s) not tagged `ai-ml`: 13 `seg-013-connect-layouts` (contact-center), 14 `seg-014-customer-profiles` (contact-center), 15 `seg-015-transform-storage` (migration), 16 `seg-016-lambda-microvms` (serverless), 19 `seg-019-devops-agent-sdk` (devops).
Use a category file (`--cats`) only when the person wants a whole section as it aired.

**Intro/outro content.** The intro is fully generic: *"This is RobotPaper's weekly AWS news podcast; the voices you're hearing are AI-generated; the stories come from public sources."* The full outro says: *"That concludes RobotPaper's weekly AWS news podcast. The recurring themes were more control over model customization, clearer operational context, and managed workflows for infrastructure changes. The source distinctions remain important: a public beta differs from general availability, and an announced future capability differs from behavior available now. The full write-up and source references are at robotpaper dot A I. From Linus, Dennis, Ken, Ada, and Grace, thank you for listening."*
For a subset build use **`seg-071-outro-short.mp3`**: the same recording with exactly this sentence cut out inside the
pauses around it (cut at 4.39 s and 14.18 s of seg-071-outro; no new speech): *"The recurring themes were more control over model customization, clearer operational context, and managed workflows for infrastructure changes."*
It keeps the plug and the credits. The credit line names all hosts of the episode even if a build uses fewer of them; keep
it (it is a recording, do not re-voice it) and say so in the show notes if needed.

## Files and format

All files: mp3, 44.1 kHz, mono, 128 kbps CBR (LAME), each well under 25 MB. **Voice only: there is no music in this kit.**

- `voice/seg-NNN-<slug>.mp3`: **one file per story item** (one announcement or one article/thread; a story that needs several
  paragraphs is still one file, e.g. `seg-068-netflix-identity`). Each is a standalone voice clip (edge-cleaned with the mixer's
  `clean_voice_edges`), laid on one voice bus and cut from its loudness-normalized (-16 LUFS) track, so all levels match.
  The first item of a category usually opens with its spoken section header, e.g. "Serverless."; the other items start straight into their story.
  The audio is the show's own voices, but the kit is not a recording of the published episode: the items are voiced one by one.
- `voice/cat-NN-<slug>.mp3`: one file per category: its items joined with 1.2 s of silence between them (the same
  gap as in segment mode). Voice only.
- `voice/seg-071-outro-short.mp3`: generic-safe outro (see above).
- `kit.json`: the JSON block at the end of this file. `build_episode.py`: the helper script below.

## Segments (one file per story item, in episode order)

| # | segment id | category | topic tags | dur (s) | host (ElevenLabs voice) | audio URL |
|---|---|---|---|---|---|---|
| 0 | `seg-000-intro` | intro  | show-intro | 8.88 | Linus (Jarnathan) | `{BASE}/voice/seg-000-intro.mp3` |
| 1 | `seg-001-bedrock-grounding` | cat-00 AI and machine learning | ai-ml | 67.04 | Ken (Liam) | `{BASE}/voice/seg-001-bedrock-grounding.mp3` |
| 2 | `seg-002-prompt-shortening` | cat-00 AI and machine learning | ai-ml | 42.56 | Linus (Jarnathan) | `{BASE}/voice/seg-002-prompt-shortening.mp3` |
| 3 | `seg-003-nemotron-on-jumpstart` | cat-00 AI and machine learning | ai-ml | 24.40 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-003-nemotron-on-jumpstart.mp3` |
| 4 | `seg-004-grok-in-govcloud` | cat-00 AI and machine learning | ai-ml | 11.76 | Linus (Jarnathan) | `{BASE}/voice/seg-004-grok-in-govcloud.mp3` |
| 5 | `seg-005-structured-outputs` | cat-00 AI and machine learning | ai-ml | 15.52 | Ken (Liam) | `{BASE}/voice/seg-005-structured-outputs.mp3` |
| 6 | `seg-006-torchneuron-beta` | cat-00 AI and machine learning | ai-ml | 71.20 | Dennis (Roger) | `{BASE}/voice/seg-006-torchneuron-beta.mp3` |
| 7 | `seg-007-sequence-lengths` | cat-00 AI and machine learning | ai-ml | 38.24 | Grace (Hope) | `{BASE}/voice/seg-007-sequence-lengths.mp3` |
| 8 | `seg-008-customization-models` | cat-00 AI and machine learning | ai-ml | 33.60 | Dennis (Roger) | `{BASE}/voice/seg-008-customization-models.mp3` |
| 9 | `seg-009-beam-on-sagemaker` | cat-00 AI and machine learning | ai-ml | 54.48 | Ken (Liam) | `{BASE}/voice/seg-009-beam-on-sagemaker.mp3` |
| 10 | `seg-010-training-session-apis` | cat-00 AI and machine learning | ai-ml | 65.74 | Linus (Jarnathan) | `{BASE}/voice/seg-010-training-session-apis.mp3` |
| 11 | `seg-011-workspaces-agent-access` | cat-00 AI and machine learning | ai-ml | 19.52 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-011-workspaces-agent-access.mp3` |
| 12 | `seg-012-agentcore-gateway` | cat-00 AI and machine learning | ai-ml | 21.04 | Grace (Hope) | `{BASE}/voice/seg-012-agentcore-gateway.mp3` |
| 13 | `seg-013-connect-layouts` | cat-00 AI and machine learning | contact-center | 30.80 | Dennis (Roger) | `{BASE}/voice/seg-013-connect-layouts.mp3` |
| 14 | `seg-014-customer-profiles` | cat-00 AI and machine learning | contact-center | 14.48 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-014-customer-profiles.mp3` |
| 15 | `seg-015-transform-storage` | cat-00 AI and machine learning | migration | 15.28 | Linus (Jarnathan) | `{BASE}/voice/seg-015-transform-storage.mp3` |
| 16 | `seg-016-lambda-microvms` | cat-00 AI and machine learning | serverless | 15.28 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-016-lambda-microvms.mp3` |
| 17 | `seg-017-openai-in-canada` | cat-00 AI and machine learning | ai-ml | 20.24 | Grace (Hope) | `{BASE}/voice/seg-017-openai-in-canada.mp3` |
| 18 | `seg-018-agentcore-sdk` | cat-00 AI and machine learning | ai-ml | 19.52 | Dennis (Roger) | `{BASE}/voice/seg-018-agentcore-sdk.mp3` |
| 19 | `seg-019-devops-agent-sdk` | cat-00 AI and machine learning | devops | 9.20 | Ken (Liam) | `{BASE}/voice/seg-019-devops-agent-sdk.mp3` |
| 20 | `seg-020-lex-sdk` | cat-00 AI and machine learning | ai-ml | 10.64 | Dennis (Roger) | `{BASE}/voice/seg-020-lex-sdk.mp3` |
| 21 | `seg-021-polly-sdk` | cat-00 AI and machine learning | ai-ml | 13.36 | Ken (Liam) | `{BASE}/voice/seg-021-polly-sdk.mp3` |
| 22 | `seg-022-knowledge-bases-sdk` | cat-00 AI and machine learning | ai-ml | 14.64 | Dennis (Roger) | `{BASE}/voice/seg-022-knowledge-bases-sdk.mp3` |
| 23 | `seg-023-lambda-durable-functions` | cat-01 Serverless | serverless | 14.24 | Ken (Liam) | `{BASE}/voice/seg-023-lambda-durable-functions.mp3` |
| 24 | `seg-024-emr-serverless-ipv6` | cat-01 Serverless | networking | 13.68 | Grace (Hope) | `{BASE}/voice/seg-024-emr-serverless-ipv6.mp3` |
| 25 | `seg-025-elasticache-migration` | cat-01 Serverless | databases | 37.20 | Linus (Jarnathan) | `{BASE}/voice/seg-025-elasticache-migration.mp3` |
| 26 | `seg-026-opensearch-refresh` | cat-01 Serverless | analytics | 14.24 | Grace (Hope) | `{BASE}/voice/seg-026-opensearch-refresh.mp3` |
| 27 | `seg-027-ec2-image-builds` | cat-02 EC2 | compute | 28.88 | Linus (Jarnathan) | `{BASE}/voice/seg-027-ec2-image-builds.mp3` |
| 28 | `seg-028-continuum-launch` | cat-03 Security Identity & Compliance | security | 62.64 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-028-continuum-launch.mp3` |
| 29 | `seg-029-identity-center-locking` | cat-03 Security Identity & Compliance | security | 17.52 | Dennis (Roger) | `{BASE}/voice/seg-029-identity-center-locking.mp3` |
| 30 | `seg-030-identity-center-sources` | cat-03 Security Identity & Compliance | security | 13.76 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-030-identity-center-sources.mp3` |
| 31 | `seg-031-guardduty-rules` | cat-03 Security Identity & Compliance | security | 13.52 | Linus (Jarnathan) | `{BASE}/voice/seg-031-guardduty-rules.mp3` |
| 32 | `seg-032-network-firewall` | cat-03 Security Identity & Compliance | security | 19.36 | Grace (Hope) | `{BASE}/voice/seg-032-network-firewall.mp3` |
| 33 | `seg-033-cognito-replication` | cat-03 Security Identity & Compliance | security | 14.88 | Ken (Liam) | `{BASE}/voice/seg-033-cognito-replication.mp3` |
| 34 | `seg-034-braket-fedramp` | cat-03 Security Identity & Compliance | security | 15.44 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-034-braket-fedramp.mp3` |
| 35 | `seg-035-rds-replica-snapshots` | cat-04 Databases | databases | 14.64 | Ken (Liam) | `{BASE}/voice/seg-035-rds-replica-snapshots.mp3` |
| 36 | `seg-036-rds-custom-updates` | cat-04 Databases | databases | 11.52 | Grace (Hope) | `{BASE}/voice/seg-036-rds-custom-updates.mp3` |
| 37 | `seg-037-dynamodb-replicas` | cat-04 Databases | databases | 16.24 | Linus (Jarnathan) | `{BASE}/voice/seg-037-dynamodb-replicas.mp3` |
| 38 | `seg-038-capacity-reservations` | cat-04 Databases | databases | 28.96 | Grace (Hope) | `{BASE}/voice/seg-038-capacity-reservations.mp3` |
| 39 | `seg-039-interconnect-last-mile` | cat-05 Networking & Content Delivery | networking | 22.56 | Linus (Jarnathan) | `{BASE}/voice/seg-039-interconnect-last-mile.mp3` |
| 40 | `seg-040-direct-connect` | cat-05 Networking & Content Delivery | networking | 13.12 | Dennis (Roger) | `{BASE}/voice/seg-040-direct-connect.mp3` |
| 41 | `seg-041-outposts-mexico` | cat-05 Networking & Content Delivery | compute | 13.12 | Ken (Liam) | `{BASE}/voice/seg-041-outposts-mexico.mp3` |
| 42 | `seg-042-log-centralization` | cat-06 Management & Governance | observability | 29.20 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-042-log-centralization.mp3` |
| 43 | `seg-043-cloudwatch-omni` | cat-06 Management & Governance | observability | 11.60 | Linus (Jarnathan) | `{BASE}/voice/seg-043-cloudwatch-omni.mp3` |
| 44 | `seg-044-alarm-tags` | cat-06 Management & Governance | observability | 20.00 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-044-alarm-tags.mp3` |
| 45 | `seg-045-ecs-gateway` | cat-07 More core services | containers | 73.04 | Dennis (Roger) | `{BASE}/voice/seg-045-ecs-gateway.mp3` |
| 46 | `seg-046-madpot-php2shell` | cat-07 More core services | security | 14.96 | Grace (Hope) | `{BASE}/voice/seg-046-madpot-php2shell.mp3` |
| 47 | `seg-047-s3-access-logs` | cat-07 More core services | storage | 20.16 | Dennis (Roger) | `{BASE}/voice/seg-047-s3-access-logs.mp3` |
| 48 | `seg-048-registration-forms` | cat-08 Applications | messaging | 13.44 | Ken (Liam) | `{BASE}/voice/seg-048-registration-forms.mp3` |
| 49 | `seg-049-rcs-wizard` | cat-08 Applications | messaging | 14.08 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-049-rcs-wizard.mp3` |
| 50 | `seg-050-brand-profiles` | cat-08 Applications | messaging | 15.60 | Ken (Liam) | `{BASE}/voice/seg-050-brand-profiles.mp3` |
| 51 | `seg-051-notify-otp` | cat-08 Applications | messaging | 19.20 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-051-notify-otp.mp3` |
| 52 | `seg-052-azure-code-conversion` | cat-09 Migration | migration | 20.00 | Linus (Jarnathan) | `{BASE}/voice/seg-052-azure-code-conversion.mp3` |
| 53 | `seg-053-landing-zones` | cat-09 Migration | migration | 16.88 | Grace (Hope) | `{BASE}/voice/seg-053-landing-zones.mp3` |
| 54 | `seg-054-network-migration` | cat-09 Migration | migration | 16.24 | Linus (Jarnathan) | `{BASE}/voice/seg-054-network-migration.mp3` |
| 55 | `seg-055-polly-voices` | cat-10 Other Announcements | ai-ml | 14.80 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-055-polly-voices.mp3` |
| 56 | `seg-056-quick-singapore` | cat-10 Other Announcements | ai-ml | 8.40 | Dennis (Roger) | `{BASE}/voice/seg-056-quick-singapore.mp3` |
| 57 | `seg-057-cloudformation-rollback` | cat-10 Other Announcements | devops | 20.72 | Ken (Liam) | `{BASE}/voice/seg-057-cloudformation-rollback.mp3` |
| 58 | `seg-058-solution-matching` | cat-11 Everything else | partner | 14.08 | Dennis (Roger) | `{BASE}/voice/seg-058-solution-matching.mp3` |
| 59 | `seg-059-program-navigator` | cat-11 Everything else | partner | 14.08 | Grace (Hope) | `{BASE}/voice/seg-059-program-navigator.mp3` |
| 60 | `seg-060-clean-rooms-databricks` | cat-11 Everything else | analytics | 13.52 | Linus (Jarnathan) | `{BASE}/voice/seg-060-clean-rooms-databricks.mp3` |
| 61 | `seg-061-global-resiliency` | cat-11 Everything else | contact-center | 17.76 | Grace (Hope) | `{BASE}/voice/seg-061-global-resiliency.mp3` |
| 62 | `seg-062-service-reference` | cat-11 Everything else | security | 11.92 | Linus (Jarnathan) | `{BASE}/voice/seg-062-service-reference.mp3` |
| 63 | `seg-063-mediaconnect-router` | cat-11 Everything else | media | 13.36 | Dennis (Roger) | `{BASE}/voice/seg-063-mediaconnect-router.mp3` |
| 64 | `seg-064-siliconangle-trainium` | cat-12 Around the web | ai-ml, community | 36.32 | Grace (Hope) | `{BASE}/voice/seg-064-siliconangle-trainium.mp3` |
| 65 | `seg-065-new-stack-pytorch` | cat-12 Around the web | ai-ml, community | 38.24 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-065-new-stack-pytorch.mp3` |
| 66 | `seg-066-repository-screening` | cat-12 Around the web | security, community | 49.12 | Ken (Liam) | `{BASE}/voice/seg-066-repository-screening.mp3` |
| 67 | `seg-067-ai-spending` | cat-12 Around the web | cost, community | 42.64 | Linus (Jarnathan) | `{BASE}/voice/seg-067-ai-spending.mp3` |
| 68 | `seg-068-netflix-identity` | cat-12 Around the web | security, community | 81.68 | Ada (Jessica Anne Bogart) | `{BASE}/voice/seg-068-netflix-identity.mp3` |
| 69 | `seg-069-security-coverage` | cat-12 Around the web | security, community | 14.48 | Dennis (Roger) | `{BASE}/voice/seg-069-security-coverage.mp3` |
| 70 | `seg-070-deployment-checks` | cat-12 Around the web | devops, community | 61.12 | Grace (Hope) | `{BASE}/voice/seg-070-deployment-checks.mp3` |
| 71 | `seg-071-outro` | outro  | show-outro | 34.56 | Linus (Jarnathan) | `{BASE}/voice/seg-071-outro.mp3` |

| 71s | `seg-071-outro-short` | outro (generic-safe edit) | show-outro | 24.78 | Linus (Jarnathan) | `{BASE}/voice/seg-071-outro-short.mp3` |

Exact file names: `seg-000-intro.mp3`, `seg-001-bedrock-grounding.mp3`, `seg-002-prompt-shortening.mp3`, `seg-003-nemotron-on-jumpstart.mp3`, `seg-004-grok-in-govcloud.mp3`, `seg-005-structured-outputs.mp3`, `seg-006-torchneuron-beta.mp3`, `seg-007-sequence-lengths.mp3`, `seg-008-customization-models.mp3`, `seg-009-beam-on-sagemaker.mp3`, `seg-010-training-session-apis.mp3`, `seg-011-workspaces-agent-access.mp3`, `seg-012-agentcore-gateway.mp3`, `seg-013-connect-layouts.mp3`, `seg-014-customer-profiles.mp3`, `seg-015-transform-storage.mp3`, `seg-016-lambda-microvms.mp3`, `seg-017-openai-in-canada.mp3`, `seg-018-agentcore-sdk.mp3`, `seg-019-devops-agent-sdk.mp3`, `seg-020-lex-sdk.mp3`, `seg-021-polly-sdk.mp3`, `seg-022-knowledge-bases-sdk.mp3`, `seg-023-lambda-durable-functions.mp3`, `seg-024-emr-serverless-ipv6.mp3`, `seg-025-elasticache-migration.mp3`, `seg-026-opensearch-refresh.mp3`, `seg-027-ec2-image-builds.mp3`, `seg-028-continuum-launch.mp3`, `seg-029-identity-center-locking.mp3`, `seg-030-identity-center-sources.mp3`, `seg-031-guardduty-rules.mp3`, `seg-032-network-firewall.mp3`, `seg-033-cognito-replication.mp3`, `seg-034-braket-fedramp.mp3`, `seg-035-rds-replica-snapshots.mp3`, `seg-036-rds-custom-updates.mp3`, `seg-037-dynamodb-replicas.mp3`, `seg-038-capacity-reservations.mp3`, `seg-039-interconnect-last-mile.mp3`, `seg-040-direct-connect.mp3`, `seg-041-outposts-mexico.mp3`, `seg-042-log-centralization.mp3`, `seg-043-cloudwatch-omni.mp3`, `seg-044-alarm-tags.mp3`, `seg-045-ecs-gateway.mp3`, `seg-046-madpot-php2shell.mp3`, `seg-047-s3-access-logs.mp3`, `seg-048-registration-forms.mp3`, `seg-049-rcs-wizard.mp3`, `seg-050-brand-profiles.mp3`, `seg-051-notify-otp.mp3`, `seg-052-azure-code-conversion.mp3`, `seg-053-landing-zones.mp3`, `seg-054-network-migration.mp3`, `seg-055-polly-voices.mp3`, `seg-056-quick-singapore.mp3`, `seg-057-cloudformation-rollback.mp3`, `seg-058-solution-matching.mp3`, `seg-059-program-navigator.mp3`, `seg-060-clean-rooms-databricks.mp3`, `seg-061-global-resiliency.mp3`, `seg-062-service-reference.mp3`, `seg-063-mediaconnect-router.mp3`, `seg-064-siliconangle-trainium.mp3`, `seg-065-new-stack-pytorch.mp3`, `seg-066-repository-screening.mp3`, `seg-067-ai-spending.mp3`, `seg-068-netflix-identity.mp3`, `seg-069-security-coverage.mp3`, `seg-070-deployment-checks.mp3`, `seg-071-outro.mp3`, `seg-071-outro-short.mp3`.

## Tags

One main topic per story item (at most two tags; a second tag is only `community`, on around-the-web items). Category tags
are the most common tag of their items. `--tags a,b` selects every item with any of those tags, in episode order.

| tag | items | segments |
|---|---|---|
| `ai-ml` | 21 | 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 17, 18, 20, 21, 22, 55, 56, 64, 65 |
| `security` | 12 | 28, 29, 30, 31, 32, 33, 34, 46, 62, 66, 68, 69 |
| `community` | 7 | 64, 65, 66, 67, 68, 69, 70 |
| `databases` | 5 | 25, 35, 36, 37, 38 |
| `messaging` | 4 | 48, 49, 50, 51 |
| `migration` | 4 | 15, 52, 53, 54 |
| `contact-center` | 3 | 13, 14, 61 |
| `devops` | 3 | 19, 57, 70 |
| `networking` | 3 | 24, 39, 40 |
| `observability` | 3 | 42, 43, 44 |
| `analytics` | 2 | 26, 60 |
| `compute` | 2 | 27, 41 |
| `partner` | 2 | 58, 59 |
| `serverless` | 2 | 16, 23 |
| `containers` | 1 | 45 |
| `cost` | 1 | 67 |
| `media` | 1 | 63 |
| `storage` | 1 | 47 |

## Categories

| category | title ("spoken section header") | segments | dur (s) | category tags | voice URL |
|---|---|---|---|---|---|
| `cat-00` | AI and machine learning ("AI and machine learning") | 1-22 | 653.74 | ai-ml | `{BASE}/voice/cat-00-ai-and-machine-learning.mp3` |
| `cat-01` | Serverless ("Serverless") | 23-26 | 82.96 | serverless | `{BASE}/voice/cat-01-serverless.mp3` |
| `cat-02` | EC2 ("Core Services, starting with compute") | 27 | 28.88 | compute | `{BASE}/voice/cat-02-ec2.mp3` |
| `cat-03` | Security Identity & Compliance ("Security, identity and compliance") | 28-34 | 164.32 | security | `{BASE}/voice/cat-03-security-identity-compliance.mp3` |
| `cat-04` | Databases ("Databases") | 35-38 | 74.96 | databases | `{BASE}/voice/cat-04-databases.mp3` |
| `cat-05` | Networking & Content Delivery ("Networking and content delivery") | 39-41 | 51.20 | networking | `{BASE}/voice/cat-05-networking-content-delivery.mp3` |
| `cat-06` | Management & Governance ("Management and governance") | 42-44 | 63.20 | observability | `{BASE}/voice/cat-06-management-governance.mp3` |
| `cat-07` | More core services ("Core Services: containers") | 45-47 | 110.56 | containers | `{BASE}/voice/cat-07-more-core-services.mp3` |
| `cat-08` | Applications ("Applications") | 48-51 | 65.92 | messaging | `{BASE}/voice/cat-08-applications.mp3` |
| `cat-09` | Migration ("Migration") | 52-54 | 55.52 | migration | `{BASE}/voice/cat-09-migration.mp3` |
| `cat-10` | Other Announcements | 55-57 | 46.32 | ai-ml | `{BASE}/voice/cat-10-other-announcements.mp3` |
| `cat-11` | Everything else | 58-63 | 90.72 | partner | `{BASE}/voice/cat-11-everything-else.mp3` |
| `cat-12` | Around the web ("Around the web and r slash AWS") | 64-70 | 330.80 | community | `{BASE}/voice/cat-12-around-the-web.mp3` |

## Stitching

All times in seconds; `prev` = end of the previous voice piece. The kit has no music, so there is only one voice bus.

```
intro voice (seg-000) at 0.3;  prev = 0.3 + 8.880
for each chosen category c (in order):
    cat voice at prev + 1.2;   prev = that start + c.dur_s
    (segment mode: the block's segment files at prev + 1.2, each next one 1.2 after the previous ends)
outro voice at prev + 1.2;  total = that start + outro dur + 0.5
mix: all voice pieces summed, loudnorm I=-16:TP=-1.5:LRA=11, mp3 128k mono 44.1k
```

Use `adelay=delays=<samples>S:all=1` with samples = round(seconds * 44100) for sample-exact placement. (The pieces carry a
little silence of their own at the start and end, so the heard gaps are a bit longer than the numbers above.)

### Worked example: the AI sections (tested)

Placements for intro + the `ai-ml` segments + short outro (total 702.122 s):

| at (s) | file |
|---|---|
| 0.300 | `voice/seg-000-intro.mp3` |
| 10.380 | `voice/seg-001-bedrock-grounding.mp3` |
| 78.620 | `voice/seg-002-prompt-shortening.mp3` |
| 122.380 | `voice/seg-003-nemotron-on-jumpstart.mp3` |
| 147.980 | `voice/seg-004-grok-in-govcloud.mp3` |
| 160.940 | `voice/seg-005-structured-outputs.mp3` |
| 177.660 | `voice/seg-006-torchneuron-beta.mp3` |
| 250.060 | `voice/seg-007-sequence-lengths.mp3` |
| 289.500 | `voice/seg-008-customization-models.mp3` |
| 324.300 | `voice/seg-009-beam-on-sagemaker.mp3` |
| 379.980 | `voice/seg-010-training-session-apis.mp3` |
| 446.922 | `voice/seg-011-workspaces-agent-access.mp3` |
| 467.642 | `voice/seg-012-agentcore-gateway.mp3` |
| 489.882 | `voice/seg-017-openai-in-canada.mp3` |
| 511.322 | `voice/seg-018-agentcore-sdk.mp3` |
| 532.042 | `voice/seg-020-lex-sdk.mp3` |
| 543.882 | `voice/seg-021-polly-sdk.mp3` |
| 558.442 | `voice/seg-022-knowledge-bases-sdk.mp3` |
| 574.282 | `voice/seg-055-polly-voices.mp3` |
| 590.282 | `voice/seg-056-quick-singapore.mp3` |
| 599.882 | `voice/seg-064-siliconangle-trainium.mp3` |
| 637.402 | `voice/seg-065-new-stack-pytorch.mp3` |
| 676.842 | `voice/seg-071-outro-short.mp3` |

```bash
# the AI sections = intro + every "ai-ml" segment (1,2,3,4,5,6,7,8,9,10,11,12,17,18,20,21,22,55,56,64,65) + generic-safe short outro. Voice only. Tested.
BASE="https://podcasts.robotpaper.ai/aws/kit/2026-10-23-week"     # the kit's {BASE}; use the base the person gave you if any
UA="Build-A-Pod-kit/1.0 (build_episode.py; +https://podcasts.robotpaper.ai/)"
mkdir -p pieces
for f in voice/seg-000-intro.mp3 voice/seg-001-bedrock-grounding.mp3 voice/seg-002-prompt-shortening.mp3 voice/seg-003-nemotron-on-jumpstart.mp3 voice/seg-004-grok-in-govcloud.mp3 voice/seg-005-structured-outputs.mp3 voice/seg-006-torchneuron-beta.mp3 voice/seg-007-sequence-lengths.mp3 voice/seg-008-customization-models.mp3 voice/seg-009-beam-on-sagemaker.mp3 voice/seg-010-training-session-apis.mp3 voice/seg-011-workspaces-agent-access.mp3 voice/seg-012-agentcore-gateway.mp3 voice/seg-017-openai-in-canada.mp3 voice/seg-018-agentcore-sdk.mp3 voice/seg-020-lex-sdk.mp3 voice/seg-021-polly-sdk.mp3 voice/seg-022-knowledge-bases-sdk.mp3 voice/seg-055-polly-voices.mp3 voice/seg-056-quick-singapore.mp3 voice/seg-064-siliconangle-trainium.mp3 voice/seg-065-new-stack-pytorch.mp3 voice/seg-071-outro-short.mp3; do
  curl -fsS -A "$UA" --create-dirs -o "pieces/$f" "$BASE/$f"
done
T=30963602                      # total length in samples = 702.122 s
ffmpeg -y -loglevel error \
  -i pieces/voice/seg-000-intro.mp3 \
  -i pieces/voice/seg-001-bedrock-grounding.mp3 \
  -i pieces/voice/seg-002-prompt-shortening.mp3 \
  -i pieces/voice/seg-003-nemotron-on-jumpstart.mp3 \
  -i pieces/voice/seg-004-grok-in-govcloud.mp3 \
  -i pieces/voice/seg-005-structured-outputs.mp3 \
  -i pieces/voice/seg-006-torchneuron-beta.mp3 \
  -i pieces/voice/seg-007-sequence-lengths.mp3 \
  -i pieces/voice/seg-008-customization-models.mp3 \
  -i pieces/voice/seg-009-beam-on-sagemaker.mp3 \
  -i pieces/voice/seg-010-training-session-apis.mp3 \
  -i pieces/voice/seg-011-workspaces-agent-access.mp3 \
  -i pieces/voice/seg-012-agentcore-gateway.mp3 \
  -i pieces/voice/seg-017-openai-in-canada.mp3 \
  -i pieces/voice/seg-018-agentcore-sdk.mp3 \
  -i pieces/voice/seg-020-lex-sdk.mp3 \
  -i pieces/voice/seg-021-polly-sdk.mp3 \
  -i pieces/voice/seg-022-knowledge-bases-sdk.mp3 \
  -i pieces/voice/seg-055-polly-voices.mp3 \
  -i pieces/voice/seg-056-quick-singapore.mp3 \
  -i pieces/voice/seg-064-siliconangle-trainium.mp3 \
  -i pieces/voice/seg-065-new-stack-pytorch.mp3 \
  -i pieces/voice/seg-071-outro-short.mp3 \
  -filter_complex "[0:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=13230S:all=1[v0];[1:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=457758S:all=1[v1];[2:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=3467142S:all=1[v2];[3:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=5396958S:all=1[v3];[4:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=6525918S:all=1[v4];[5:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=7097454S:all=1[v5];[6:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=7834806S:all=1[v6];[7:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=11027646S:all=1[v7];[8:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=12766950S:all=1[v8];[9:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=14301630S:all=1[v9];[10:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=16757118S:all=1[v10];[11:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=19709282S:all=1[v11];[12:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=20623034S:all=1[v12];[13:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=21603818S:all=1[v13];[14:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=22549322S:all=1[v14];[15:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=23463074S:all=1[v15];[16:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=23985218S:all=1[v16];[17:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=24627314S:all=1[v17];[18:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=25325858S:all=1[v18];[19:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=26031458S:all=1[v19];[20:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=26454818S:all=1[v20];[21:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=28109450S:all=1[v21];[22:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays=29848754S:all=1[v22];[v0][v1][v2][v3][v4][v5][v6][v7][v8][v9][v10][v11][v12][v13][v14][v15][v16][v17][v18][v19][v20][v21][v22]amix=inputs=23:normalize=0:duration=longest,apad=whole_len=$T,atrim=end_sample=$T,loudnorm=I=-16:TP=-1.5:LRA=11,aresample=44100[o]" \
  -map '[o]' -ar 44100 -ac 1 -codec:a libmp3lame -b:a 128k ai-sections-only.mp3
```

### Subsets smaller than a category

Use the per-segment files in a run: the segments of one category that follow each other are placed 1.2 s
apart, and the run starts 1.2 s after the previous block. The helper does all of this: `--segs 64,65,66`.
Each category's first item usually opens with its spoken section header (e.g. segment 1 starts "AI and machine learning.");
the other items start straight into their story. Segments are in episode order; pick any items you like, one per story.

### Helper script (`build_episode.py`, stdlib only; also at `{BASE}/build_episode.py`)

Save the JSON block at the end of this file as `kit.json` (or download `{BASE}/kit.json`) and this script as
`build_episode.py`. It downloads only the files the build needs that are missing from `--dir`, from kit.json
`base_url` (= BASE above) or from `--base URL` if given (it sends a User-Agent header); `--local` never downloads (use a
complete local kit folder). It checks every piece against the `bytes` and `md5` in kit.json (re-downloads a cached piece
that does not match, and stops if a download still does not match: the CDN may be serving a stale copy). `--tags`,
`--segs` and `--cats` are sorted and de-duplicated.

```python
#!/usr/bin/env python3
"""Stitch a custom cut of an AWS weekly episode from the pre-rendered kit voice pieces (stdlib + ffmpeg only; no TTS,
no API calls). The kit is VOICE ONLY: it contains no music, and this script adds none.

  python3 build_episode.py --kit kit.json --dir pieces/ --cats 0 --outro short -o ai-sections-only.mp3
      downloads the pieces it needs (only missing ones) from kit.json "base_url" (the public kit) into --dir
  --base URL        download from another base instead      --local      never download, use --dir as is
  --cats 0,3        category mode (one whole-category file each)
  --segs 3,4,5      segment mode (pieces of categories; consecutive segments of one category form one block)
  --tags cost,security   segment mode with every story segment that has any of these tags
  --print           print the ffmpeg command, do not run it
Give exactly one of --cats, --segs, --tags. Ids are sorted and de-duplicated (with a note on stderr); unknown ids,
unknown tags and bad lists exit with a message. Every piece used (downloaded, reused from --dir, or --local) is checked
against kit.json "bytes" and "md5": a reused file that differs is downloaded again once; a download that still differs,
or a --local file that is missing or differs, is an error. Kits without bytes/md5 skip the check (with a note).
Timeline: intro voice at timing.intro_lead_s; every chosen block (a category file, or a run of segment files) starts
timing.block_gap_s after the previous voice piece ends; segments of one block follow each other with
timing.segment_gap_s; the outro starts timing.block_gap_s after the last block; the file ends timing.tail_s after the
outro. All pieces are mixed on one voice bus, then loudnorm I=-16 TP=-1.5 LRA=11, 44.1 kHz mono 128k."""
import argparse, hashlib, json, os, shlex, subprocess, sys, urllib.request
# Fallback only, used when kit.json has no base_url and no --base is given. It is episode-specific (2026-10-23-week):
# every kit made by make_md.py carries its own base_url, so this only matters for hand-made kit.json files.
DEFAULT_BASE = "https://podcasts.robotpaper.ai/aws/kit/2026-10-23-week"
UA = "Build-A-Pod-kit/1.0 (build_episode.py; +https://podcasts.robotpaper.ai/)"   # the site blocks Python-urllib's UA

SR = 44100
NORM = "loudnorm=I=-16:TP=-1.5:LRA=11"

def note(msg): print(f"note: {msg}", file=sys.stderr)

def span(ids):
    """[1,2,3,5] -> '1-3, 5' (for error messages)."""
    out, ids = [], sorted(ids)
    for i in ids:
        if out and i == out[-1][1] + 1: out[-1][1] = i
        else: out.append([i, i])
    return ", ".join(f"{a}" if a == b else f"{a}-{b}" for a, b in out)

def valid_ids(kit, what):
    if what == "cats": return [c["index"] for c in kit["categories"]]
    return [s["order"] for s in kit["segments"] if s["category"] is not None]

def parse_ids(kit, what, raw):
    """'5,3,3' -> [3, 5]: checks syntax and ids, sorts and de-duplicates (and says so on stderr)."""
    try: ids = [int(x) for x in raw.split(",")]
    except ValueError: raise SystemExit(f"--{what} {raw!r}: give a comma-separated list of integers, e.g. --{what} 3,4,5")
    ok, byseg = valid_ids(kit, what), {s["order"]: s for s in kit["segments"]}
    for i in ids:
        if what == "segs" and i in byseg and byseg[i]["category"] is None:
            raise SystemExit(f"segment {i} is intro/outro; they are added automatically")
        if i not in ok: raise SystemExit(f"--{what}: unknown id {i}; valid ids are {span(ok)}")
    s = sorted(set(ids))
    if s != ids: note(f"--{what} {raw} sorted and de-duplicated to {','.join(map(str, s))} (episode order, each once)")
    return s

def tag_segs(kit, raw):
    """'cost,security' -> sorted story segment orders that carry any of those tags."""
    want = [x.strip() for x in raw.split(",")]
    have = sorted({g for s in kit["segments"] if s["category"] is not None for g in s.get("tags", [])})
    bad = [x for x in want if x not in have]
    if bad: raise SystemExit(f"--tags: unknown tag(s) {', '.join(map(repr, bad))}; tags in this kit: {', '.join(have)}")
    segs = sorted(s["order"] for s in kit["segments"] if s["category"] is not None and set(want) & set(s.get("tags", [])))
    note(f"--tags {raw} -> --segs {','.join(map(str, segs))}")
    return segs

def plan(kit, cats=None, segs=None, outro="full"):
    """List of events (file, at_s) and the total length T (s). Voice pieces only."""
    t = kit["timing"]
    if outro != "full" and "outro_short" not in kit: raise SystemExit("this kit has no short outro; use --outro full")
    intro, out = kit["intro"], kit["outro"] if outro == "full" else kit["outro_short"]
    ev = [(intro["voice"], t["intro_lead_s"])]
    prev = t["intro_lead_s"] + intro["dur_s"]
    bycat = {c["index"]: c for c in kit["categories"]}
    byseg = {s["order"]: s for s in kit["segments"]}
    for i in cats or []:
        if i not in bycat: raise SystemExit(f"unknown category {i}; valid ids are {span(bycat)}")
    for i in segs or []:
        if i not in byseg: raise SystemExit(f"unknown segment {i}; valid ids are {span(valid_ids(kit, 'segs'))}")
    blocks = []
    if cats is not None: blocks = [("cat", bycat[c], None) for c in cats]
    else:   # consecutive segments of one category -> one block
        for o in segs:
            s = byseg[o]
            if s["category"] is None: raise SystemExit(f"segment {o} is intro/outro; they are added automatically")
            if blocks and blocks[-1][1]["index"] == s["category"] and blocks[-1][2][-1] == o - 1: blocks[-1][2].append(o)
            else: blocks.append(("seg", bycat[s["category"]], [o]))
    for mode, c, ss in blocks:
        v0 = prev + t["block_gap_s"]
        if mode == "cat" or ss == c["segments"]:   # whole category: its own file (segments joined as in the episode)
            ev.append((c["voice"], v0)); prev = v0 + c["dur_s"]; continue
        x = v0
        for o in ss:
            s = byseg[o]
            if o != ss[0]: x += t["segment_gap_s"]
            ev.append((s["file"], x)); x += s["dur_s"]
        prev = x
    ev.append((out["voice"], prev + t["block_gap_s"]))
    return ev, prev + t["block_gap_s"] + out["dur_s"] + t["tail_s"]

def ffmpeg_cmd(ev, T, d, out):
    args, fv = ["ffmpeg", "-y", "-loglevel", "error"], []
    for n, (f, at) in enumerate(ev):
        args += ["-i", os.path.join(d, f)]
        fv.append(f"[{n}:a]aresample=44100,aformat=channel_layouts=mono,adelay=delays={round(at * SR)}S:all=1[v{n}]")
    pad = f"apad=whole_len={round(T * SR)},atrim=end_sample={round(T * SR)}"
    g = fv + [f"{''.join(f'[v{n}]' for n in range(len(ev)))}amix=inputs={len(ev)}:normalize=0:duration=longest,{pad},{NORM},aresample=44100[o]"]
    return args + ["-filter_complex", ";".join(g), "-map", "[o]", "-ar", "44100", "-ac", "1",
                   "-codec:a", "libmp3lame", "-b:a", "128k", out]

def expected(kit):
    """path -> (bytes, md5) for every file-bearing kit entry that has both fields (old kits: empty)."""
    es = [kit["intro"], kit["outro"]] + ([kit["outro_short"]] if "outro_short" in kit else []) + kit["categories"] + kit["segments"]
    return {e.get("voice") or e["file"]: (e["bytes"], e["md5"]) for e in es if "bytes" in e and "md5" in e}

def mismatch(path, want):
    """None if the file at path has the expected size and md5, else a short reason."""
    if not os.path.exists(path): return "missing"
    n = os.path.getsize(path)
    if n != want[0]: return f"{n} bytes, kit.json says {want[0]}"
    h = hashlib.md5()
    with open(path, "rb") as f:
        for chunk in iter(lambda: f.read(1 << 20), b""): h.update(chunk)
    return None if h.hexdigest() == want[1] else f"md5 {h.hexdigest()}, kit.json says {want[1]}"

def fetch(url, path, want=None):
    """Download url to path; with want=(bytes, md5) a download that does not match is an error (file not kept)."""
    req = urllib.request.Request(url, headers={"User-Agent": UA})
    with urllib.request.urlopen(req, timeout=60) as r, open(path + ".part", "wb") as f:
        while True:
            chunk = r.read(1 << 16)
            if not chunk: break
            f.write(chunk)
    bad = want and mismatch(path + ".part", want)
    if bad:
        os.remove(path + ".part")
        raise SystemExit(f"downloaded {url} does not match kit.json ({bad}); the CDN may be serving a stale cached "
                         "copy: purge it or retry later")
    os.replace(path + ".part", path)

def get_files(files, d, base, exp):
    """Make sure every file is in d and matches kit.json (download missing/mismatching ones unless base is None)."""
    if not exp: note("kit.json has no bytes/md5 fields (old kit); pieces are not verified")
    elif any(f not in exp for f in files):
        note(f"no bytes/md5 in kit.json for {', '.join(f for f in files if f not in exp)}; not verified")
    for f in files:
        p, want = os.path.join(d, f), exp.get(f)
        bad = mismatch(p, want) if want else None if os.path.exists(p) else "missing"
        if not bad: continue
        if base is None: raise SystemExit(f"--local: {p} is missing" if bad == "missing"
                                          else f"--local: {p} does not match kit.json ({bad})")
        if bad != "missing": note(f"{p} does not match kit.json ({bad}); downloading it again")
        os.makedirs(os.path.dirname(p) or ".", exist_ok=True)
        fetch(base.rstrip("/") + "/" + f, p, want)

def main():
    ap = argparse.ArgumentParser()
    ap.add_argument("--kit", default="kit.json"); ap.add_argument("--dir", default=".")
    ap.add_argument("--base"); ap.add_argument("--local", action="store_true")
    ap.add_argument("--cats"); ap.add_argument("--segs"); ap.add_argument("--tags")
    ap.add_argument("--outro", choices=["full", "short"], default="full")
    ap.add_argument("--print", action="store_true")
    ap.add_argument("-o", "--out", default="episode.mp3")
    a = ap.parse_args()
    with open(a.kit) as f: kit = json.load(f)
    if sum(x is not None for x in (a.cats, a.segs, a.tags)) != 1:
        raise SystemExit("give exactly one of --cats, --segs or --tags")
    cats = parse_ids(kit, "cats", a.cats) if a.cats is not None else None
    segs = parse_ids(kit, "segs", a.segs) if a.segs is not None else tag_segs(kit, a.tags) if a.tags is not None else None
    ev, T = plan(kit, cats=cats, segs=segs, outro=a.outro)
    base = None if a.local else a.base or kit.get("base_url") or DEFAULT_BASE
    get_files(sorted({e[0] for e in ev}), a.dir, base, expected(kit))
    cmd = ffmpeg_cmd(ev, T, a.dir, a.out)
    for f, at in sorted(ev, key=lambda e: e[1]): print(f"{at:9.3f}s  voice  {f}")
    print(f"total {T:.3f}s")
    if a.print: print(" ".join(shlex.quote(x) for x in cmd)); return
    subprocess.run(cmd, check=True)
    print(f"wrote {a.out} (voice only; the kit has no music)")

if __name__ == "__main__":
    main()
```

## Verify (do all of these before handing the mp3 over)

```bash
OUT=ai-sections-only.mp3
ffprobe -v error -show_entries format=duration:stream=sample_rate,channels,bit_rate -of default=nw=1 $OUT
#   expect duration ~ 702.1 s (+ ~0.025-0.05 s mp3 padding), 44100 Hz, 1 channel, 128000
ffmpeg -nostats -i $OUT -af ebur128=peak=true -f null - 2>&1 | sed -n '/Summary/,$p' | grep -E 'I:|Peak:'
#   expect I about -16 LUFS (accept -18 to -15; loudnorm is single-pass), true peak about -1.5 dBFS; accept <= -1.0
ffmpeg -nostats -i $OUT -af silencedetect=noise=-45dB:d=0.9 -f null - 2>&1 | grep -o 'silence_start: [0-9.]*\|silence_duration: [0-9.]*' | paste - -
#   expect only planned gaps, each a bit longer than listed because the clips' own head/tail silence is included:
#   about 1.3-1.7 s between pieces
```

Also listen to (or RMS-check) the first 3 s of every voice piece you placed: speech must start where planned and no
piece may end before its listed duration (no truncation).

## Machine-readable data (`kit.json`)

```json
{
 "kit": "Build-A-Pod prototype kit (voice only, one file per story item)",
 "show": "RobotPaper weekly AWS news podcast",
 "episode": "2026-10-23-week",
 "format": {
  "codec": "mp3",
  "sample_rate": 44100,
  "channels": 1,
  "bitrate": "128k",
  "loudness_target_lufs": -16,
  "true_peak_db": -1.5,
  "lra": 11
 },
 "base_url": "https://podcasts.robotpaper.ai/aws/kit/2026-10-23-week",
 "base_url_placeholder": "{BASE}",
 "timing": {
  "intro_lead_s": 0.3,
  "block_gap_s": 1.2,
  "segment_gap_s": 1.2,
  "tail_s": 0.5,
  "note": "start times of the voice pieces: intro at intro_lead_s; each block (category file or run of segment files) block_gap_s after the previous voice piece ends; segments inside a block segment_gap_s apart; outro block_gap_s after the last block; file ends tail_s after the outro. Gaps inside a cat-* file are baked in (1.2 s, same as segment_gap_s). One segment file = one story item."
 },
 "intro": {
  "voice": "voice/seg-000-intro.mp3",
  "dur_s": 8.88,
  "text": "This is RobotPaper's weekly AWS news podcast; the voices you're hearing are AI-generated; the stories come from public sources.",
  "bytes": 142985,
  "md5": "3dff19ea54a208df7f947bbc2758ee6b"
 },
 "outro": {
  "voice": "voice/seg-071-outro.mp3",
  "dur_s": 34.56,
  "text": "That concludes RobotPaper's weekly AWS news podcast. The recurring themes were more control over model customization, clearer operational context, and managed workflows for infrastructure changes. The source distinctions remain important: a public beta differs from general availability, and an announced future capability differs from behavior available now. The full write-up and source references are at robotpaper dot A I. From Linus, Dennis, Ken, Ada, and Grace, thank you for listening.",
  "bytes": 553838,
  "md5": "3d8973600f3a7b04fa81e42cf781052a"
 },
 "outro_short": {
  "voice": "voice/seg-071-outro-short.mp3",
  "dur_s": 24.78,
  "cut_s": [
   4.39,
   14.18
  ],
  "text": "That concludes RobotPaper's weekly AWS news podcast. The source distinctions remain important: a public beta differs from general availability, and an announced future capability differs from behavior available now. The full write-up and source references are at robotpaper dot A I. From Linus, Dennis, Ken, Ada, and Grace, thank you for listening.",
  "note": "the outro with the episode-wide sentence cut out (edit of the existing audio, no TTS)",
  "bytes": 397522,
  "md5": "5cca315dbe161062e493771504417034"
 },
 "categories": [
  {
   "index": 0,
   "id": "cat-00",
   "title": "AI and machine learning",
   "section": "AI and machine learning",
   "tags": [
    "ai-ml"
   ],
   "segments": [
    1,
    2,
    3,
    4,
    5,
    6,
    7,
    8,
    9,
    10,
    11,
    12,
    13,
    14,
    15,
    16,
    17,
    18,
    19,
    20,
    21,
    22
   ],
   "voice": "voice/cat-00-ai-and-machine-learning.mp3",
   "dur_s": 653.742494,
   "bytes": 10461143,
   "md5": "b83c1aae881c72ab51a2a22cf54649cf"
  },
  {
   "index": 1,
   "id": "cat-01",
   "title": "Serverless",
   "section": "Serverless",
   "tags": [
    "serverless"
   ],
   "segments": [
    23,
    24,
    25,
    26
   ],
   "voice": "voice/cat-01-serverless.mp3",
   "dur_s": 82.96,
   "bytes": 1328317,
   "md5": "b7593bff0f6f844c26eea87fd2c0adfd"
  },
  {
   "index": 2,
   "id": "cat-02",
   "title": "EC2",
   "section": "Core Services, starting with compute",
   "tags": [
    "compute"
   ],
   "segments": [
    27
   ],
   "voice": "voice/cat-02-ec2.mp3",
   "dur_s": 28.88,
   "bytes": 463141,
   "md5": "15be576507f38d91e2d1d3209a488438"
  },
  {
   "index": 3,
   "id": "cat-03",
   "title": "Security Identity & Compliance",
   "section": "Security, identity and compliance",
   "tags": [
    "security"
   ],
   "segments": [
    28,
    29,
    30,
    31,
    32,
    33,
    34
   ],
   "voice": "voice/cat-03-security-identity-compliance.mp3",
   "dur_s": 164.32,
   "bytes": 2630260,
   "md5": "6a75de5c97d1002b9c5454348a5ae6fc"
  },
  {
   "index": 4,
   "id": "cat-04",
   "title": "Databases",
   "section": "Databases",
   "tags": [
    "databases"
   ],
   "segments": [
    35,
    36,
    37,
    38
   ],
   "voice": "voice/cat-04-databases.mp3",
   "dur_s": 74.96,
   "bytes": 1200421,
   "md5": "316d432a8a6ac0d6609d1b5139a3097b"
  },
  {
   "index": 5,
   "id": "cat-05",
   "title": "Networking & Content Delivery",
   "section": "Networking and content delivery",
   "tags": [
    "networking"
   ],
   "segments": [
    39,
    40,
    41
   ],
   "voice": "voice/cat-05-networking-content-delivery.mp3",
   "dur_s": 51.2,
   "bytes": 820078,
   "md5": "b5db75c7c89b02bbc692aea9f0198fd4"
  },
  {
   "index": 6,
   "id": "cat-06",
   "title": "Management & Governance",
   "section": "Management and governance",
   "tags": [
    "observability"
   ],
   "segments": [
    42,
    43,
    44
   ],
   "voice": "voice/cat-06-management-governance.mp3",
   "dur_s": 63.2,
   "bytes": 1012340,
   "md5": "19c205250b5c8a87297c29b05b174782"
  },
  {
   "index": 7,
   "id": "cat-07",
   "title": "More core services",
   "section": "Core Services: containers",
   "tags": [
    "containers"
   ],
   "segments": [
    45,
    46,
    47
   ],
   "voice": "voice/cat-07-more-core-services.mp3",
   "dur_s": 110.56,
   "bytes": 1770100,
   "md5": "867f1f980d4d341834dbeb02b216c93a"
  },
  {
   "index": 8,
   "id": "cat-08",
   "title": "Applications",
   "section": "Applications",
   "tags": [
    "messaging"
   ],
   "segments": [
    48,
    49,
    50,
    51
   ],
   "voice": "voice/cat-08-applications.mp3",
   "dur_s": 65.92,
   "bytes": 1055807,
   "md5": "6790e4aff22e3a66fa5cee1e88930cce"
  },
  {
   "index": 9,
   "id": "cat-09",
   "title": "Migration",
   "section": "Migration",
   "tags": [
    "migration"
   ],
   "segments": [
    52,
    53,
    54
   ],
   "voice": "voice/cat-09-migration.mp3",
   "dur_s": 55.52,
   "bytes": 889460,
   "md5": "94406e452d77d8269a98ed3a90b770aa"
  },
  {
   "index": 10,
   "id": "cat-10",
   "title": "Other Announcements",
   "section": "",
   "tags": [
    "ai-ml"
   ],
   "segments": [
    55,
    56,
    57
   ],
   "voice": "voice/cat-10-other-announcements.mp3",
   "dur_s": 46.32,
   "bytes": 742338,
   "md5": "d30be39982e8b75ec607de1c79e9ebd6"
  },
  {
   "index": 11,
   "id": "cat-11",
   "title": "Everything else",
   "section": "",
   "tags": [
    "partner"
   ],
   "segments": [
    58,
    59,
    60,
    61,
    62,
    63
   ],
   "voice": "voice/cat-11-everything-else.mp3",
   "dur_s": 90.72,
   "bytes": 1452451,
   "md5": "9beffc879dc539a7ecef31b9b5097e58"
  },
  {
   "index": 12,
   "id": "cat-12",
   "title": "Around the web",
   "section": "Around the web and r slash AWS",
   "tags": [
    "community"
   ],
   "segments": [
    64,
    65,
    66,
    67,
    68,
    69,
    70
   ],
   "voice": "voice/cat-12-around-the-web.mp3",
   "dur_s": 330.8,
   "bytes": 5293914,
   "md5": "fc6e8554eeafdacc33482f69a96bb428"
  }
 ],
 "segments": [
  {
   "order": 0,
   "id": "seg-000-intro",
   "file": "voice/seg-000-intro.mp3",
   "kind": "intro",
   "category": null,
   "chapter": "Intro",
   "tags": [
    "show-intro"
   ],
   "dur_s": 8.88,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 142985,
   "md5": "3dff19ea54a208df7f947bbc2758ee6b"
  },
  {
   "order": 1,
   "id": "seg-001-bedrock-grounding",
   "file": "voice/seg-001-bedrock-grounding.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Bedrock grounding",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 67.04,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 1073780,
   "md5": "d0e361f313bc12d94a8b334bf20584e3"
  },
  {
   "order": 2,
   "id": "seg-002-prompt-shortening",
   "file": "voice/seg-002-prompt-shortening.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Prompt shortening",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 42.56,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 682152,
   "md5": "1e90b1f0c859109f1470f45e21234ae8"
  },
  {
   "order": 3,
   "id": "seg-003-nemotron-on-jumpstart",
   "file": "voice/seg-003-nemotron-on-jumpstart.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Nemotron on JumpStart",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 24.4,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 391670,
   "md5": "f4935eaa0b7eb7751f06af660620766b"
  },
  {
   "order": 4,
   "id": "seg-004-grok-in-govcloud",
   "file": "voice/seg-004-grok-in-govcloud.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Grok in GovCloud",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 11.76,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 189378,
   "md5": "5bd39b6e83894b4be356f2e05b44c902"
  },
  {
   "order": 5,
   "id": "seg-005-structured-outputs",
   "file": "voice/seg-005-structured-outputs.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Structured outputs",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 15.52,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 249564,
   "md5": "62e1b43e1b66c7bc56a754d69891849a"
  },
  {
   "order": 6,
   "id": "seg-006-torchneuron-beta",
   "file": "voice/seg-006-torchneuron-beta.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "TorchNeuron beta",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 71.2,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 1140235,
   "md5": "9cb4c83e2634779b7e7137c4c19f4fbb"
  },
  {
   "order": 7,
   "id": "seg-007-sequence-lengths",
   "file": "voice/seg-007-sequence-lengths.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Sequence lengths",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 38.24,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 612771,
   "md5": "562aa12c8c544be06e218f1c28c6a1c9"
  },
  {
   "order": 8,
   "id": "seg-008-customization-models",
   "file": "voice/seg-008-customization-models.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Customization models",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 33.6,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 538792,
   "md5": "39058e8fb2eb66ef1104a736a5ffeac8"
  },
  {
   "order": 9,
   "id": "seg-009-beam-on-sagemaker",
   "file": "voice/seg-009-beam-on-sagemaker.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Beam on SageMaker",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 54.48,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 872741,
   "md5": "aec0c52974149987186d04d656208b2e"
  },
  {
   "order": 10,
   "id": "seg-010-training-session-apis",
   "file": "voice/seg-010-training-session-apis.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Training session APIs",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 65.742494,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 1052882,
   "md5": "ba721d7dfcae2cbabdf21b6f3cbdc139"
  },
  {
   "order": 11,
   "id": "seg-011-workspaces-agent-access",
   "file": "voice/seg-011-workspaces-agent-access.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "WorkSpaces agent access",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 19.52,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 313512,
   "md5": "9301e50427e979e7ecf57399b1449471"
  },
  {
   "order": 12,
   "id": "seg-012-agentcore-gateway",
   "file": "voice/seg-012-agentcore-gateway.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "AgentCore Gateway",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 21.04,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 337754,
   "md5": "9fece58aa6aa644f03304b8edef4730e"
  },
  {
   "order": 13,
   "id": "seg-013-connect-layouts",
   "file": "voice/seg-013-connect-layouts.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Connect layouts",
   "tags": [
    "contact-center"
   ],
   "dur_s": 30.8,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 494070,
   "md5": "377015c5b3f375974ec102b05be3a40f"
  },
  {
   "order": 14,
   "id": "seg-014-customer-profiles",
   "file": "voice/seg-014-customer-profiles.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Customer Profiles",
   "tags": [
    "contact-center"
   ],
   "dur_s": 14.48,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 232846,
   "md5": "fda4053a23599769d469fd8ca955c251"
  },
  {
   "order": 15,
   "id": "seg-015-transform-storage",
   "file": "voice/seg-015-transform-storage.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Transform storage",
   "tags": [
    "migration"
   ],
   "dur_s": 15.28,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 245385,
   "md5": "eb50ab9f3dcf1f88921441d4ca05b124"
  },
  {
   "order": 16,
   "id": "seg-016-lambda-microvms",
   "file": "voice/seg-016-lambda-microvms.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Lambda MicroVMs",
   "tags": [
    "serverless"
   ],
   "dur_s": 15.28,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 245385,
   "md5": "9c0ad422c49f267e0a74a5564a8ed09d"
  },
  {
   "order": 17,
   "id": "seg-017-openai-in-canada",
   "file": "voice/seg-017-openai-in-canada.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "OpenAI in Canada",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 20.24,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 324797,
   "md5": "99f4e744e13b9e44ae4acb410244e376"
  },
  {
   "order": 18,
   "id": "seg-018-agentcore-sdk",
   "file": "voice/seg-018-agentcore-sdk.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "AgentCore SDK",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 19.52,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 313512,
   "md5": "dd679c77cbdf08260309383637c8465a"
  },
  {
   "order": 19,
   "id": "seg-019-devops-agent-sdk",
   "file": "voice/seg-019-devops-agent-sdk.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "DevOps Agent SDK",
   "tags": [
    "devops"
   ],
   "dur_s": 9.2,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 148418,
   "md5": "3d63c7a7ebf5b38022ec6960e08c2d46"
  },
  {
   "order": 20,
   "id": "seg-020-lex-sdk",
   "file": "voice/seg-020-lex-sdk.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Lex SDK",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 10.64,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 171406,
   "md5": "28c64a0a7a2167ac9bfe0f65ac29b7b0"
  },
  {
   "order": 21,
   "id": "seg-021-polly-sdk",
   "file": "voice/seg-021-polly-sdk.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Polly SDK",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 13.36,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 214874,
   "md5": "50eaa7aa04b469926df0803c22d33bce"
  },
  {
   "order": 22,
   "id": "seg-022-knowledge-bases-sdk",
   "file": "voice/seg-022-knowledge-bases-sdk.mp3",
   "kind": "story",
   "category": 0,
   "chapter": "Knowledge Bases SDK",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 14.64,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 235354,
   "md5": "4729425f72ed183fa0ae3579193a4b4b"
  },
  {
   "order": 23,
   "id": "seg-023-lambda-durable-functions",
   "file": "voice/seg-023-lambda-durable-functions.mp3",
   "kind": "story",
   "category": 1,
   "chapter": "Lambda durable functions",
   "tags": [
    "serverless"
   ],
   "dur_s": 14.24,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 229084,
   "md5": "2201b3d32e2dda49466f2c8d10ca23d9"
  },
  {
   "order": 24,
   "id": "seg-024-emr-serverless-ipv6",
   "file": "voice/seg-024-emr-serverless-ipv6.mp3",
   "kind": "story",
   "category": 1,
   "chapter": "EMR Serverless IPv6",
   "tags": [
    "networking"
   ],
   "dur_s": 13.68,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 219889,
   "md5": "a16abd485a22bf061022e5fe60638405"
  },
  {
   "order": 25,
   "id": "seg-025-elasticache-migration",
   "file": "voice/seg-025-elasticache-migration.mp3",
   "kind": "story",
   "category": 1,
   "chapter": "ElastiCache migration",
   "tags": [
    "databases"
   ],
   "dur_s": 37.2,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 596470,
   "md5": "0fd1ccdaf4b2af92f2fd1bd4c06a7216"
  },
  {
   "order": 26,
   "id": "seg-026-opensearch-refresh",
   "file": "voice/seg-026-opensearch-refresh.mp3",
   "kind": "story",
   "category": 1,
   "chapter": "OpenSearch refresh",
   "tags": [
    "analytics"
   ],
   "dur_s": 14.24,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 229084,
   "md5": "b6d1207443bdad654b5920cae11f265d"
  },
  {
   "order": 27,
   "id": "seg-027-ec2-image-builds",
   "file": "voice/seg-027-ec2-image-builds.mp3",
   "kind": "story",
   "category": 2,
   "chapter": "EC2 image builds",
   "tags": [
    "compute"
   ],
   "dur_s": 28.88,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 463141,
   "md5": "15be576507f38d91e2d1d3209a488438"
  },
  {
   "order": 28,
   "id": "seg-028-continuum-launch",
   "file": "voice/seg-028-continuum-launch.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Continuum launch",
   "tags": [
    "security"
   ],
   "dur_s": 62.64,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 1003145,
   "md5": "c5aed2af703cd8bef364d8811bbda0b3"
  },
  {
   "order": 29,
   "id": "seg-029-identity-center-locking",
   "file": "voice/seg-029-identity-center-locking.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Identity Center locking",
   "tags": [
    "security"
   ],
   "dur_s": 17.52,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 281329,
   "md5": "dd2850b4699f155693979c504d709be8"
  },
  {
   "order": 30,
   "id": "seg-030-identity-center-sources",
   "file": "voice/seg-030-identity-center-sources.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Identity Center sources",
   "tags": [
    "security"
   ],
   "dur_s": 13.76,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 221143,
   "md5": "e6bff696a5e331c633a114b3412f2fae"
  },
  {
   "order": 31,
   "id": "seg-031-guardduty-rules",
   "file": "voice/seg-031-guardduty-rules.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "GuardDuty rules",
   "tags": [
    "security"
   ],
   "dur_s": 13.52,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 217381,
   "md5": "84fbfd1f1d5fa9da0987ee72df96b357"
  },
  {
   "order": 32,
   "id": "seg-032-network-firewall",
   "file": "voice/seg-032-network-firewall.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Network Firewall",
   "tags": [
    "security"
   ],
   "dur_s": 19.36,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 311004,
   "md5": "54d2218778e7b1854942bcc2f8c840da"
  },
  {
   "order": 33,
   "id": "seg-033-cognito-replication",
   "file": "voice/seg-033-cognito-replication.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Cognito replication",
   "tags": [
    "security"
   ],
   "dur_s": 14.88,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 239115,
   "md5": "815ae810773c4223c530ac5690d9197d"
  },
  {
   "order": 34,
   "id": "seg-034-braket-fedramp",
   "file": "voice/seg-034-braket-fedramp.mp3",
   "kind": "story",
   "category": 3,
   "chapter": "Braket FedRAMP",
   "tags": [
    "security"
   ],
   "dur_s": 15.44,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 248310,
   "md5": "7490557cf8f20fb7a79cf8e16488f403"
  },
  {
   "order": 35,
   "id": "seg-035-rds-replica-snapshots",
   "file": "voice/seg-035-rds-replica-snapshots.mp3",
   "kind": "story",
   "category": 4,
   "chapter": "RDS replica snapshots",
   "tags": [
    "databases"
   ],
   "dur_s": 14.64,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 235354,
   "md5": "3c459d0b79c55eb896b9bf2197dad950"
  },
  {
   "order": 36,
   "id": "seg-036-rds-custom-updates",
   "file": "voice/seg-036-rds-custom-updates.mp3",
   "kind": "story",
   "category": 4,
   "chapter": "RDS Custom updates",
   "tags": [
    "databases"
   ],
   "dur_s": 11.52,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 185198,
   "md5": "825ff6b15a55c28eb94b3d3844f256a2"
  },
  {
   "order": 37,
   "id": "seg-037-dynamodb-replicas",
   "file": "voice/seg-037-dynamodb-replicas.mp3",
   "kind": "story",
   "category": 4,
   "chapter": "DynamoDB replicas",
   "tags": [
    "databases"
   ],
   "dur_s": 16.24,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 260849,
   "md5": "cef145c0443b20b9591ca7200e6ada5c"
  },
  {
   "order": 38,
   "id": "seg-038-capacity-reservations",
   "file": "voice/seg-038-capacity-reservations.mp3",
   "kind": "story",
   "category": 4,
   "chapter": "Capacity reservations",
   "tags": [
    "databases"
   ],
   "dur_s": 28.96,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 464395,
   "md5": "77c6884f2693d846662a535740dafaad"
  },
  {
   "order": 39,
   "id": "seg-039-interconnect-last-mile",
   "file": "voice/seg-039-interconnect-last-mile.mp3",
   "kind": "story",
   "category": 5,
   "chapter": "Interconnect last mile",
   "tags": [
    "networking"
   ],
   "dur_s": 22.56,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 361995,
   "md5": "d6b47ecad87eb307b080f433296cd148"
  },
  {
   "order": 40,
   "id": "seg-040-direct-connect",
   "file": "voice/seg-040-direct-connect.mp3",
   "kind": "story",
   "category": 5,
   "chapter": "Direct Connect",
   "tags": [
    "networking"
   ],
   "dur_s": 13.12,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 211112,
   "md5": "f5cf30e9e25b5334abd3876ff962295f"
  },
  {
   "order": 41,
   "id": "seg-041-outposts-mexico",
   "file": "voice/seg-041-outposts-mexico.mp3",
   "kind": "story",
   "category": 5,
   "chapter": "Outposts Mexico",
   "tags": [
    "compute"
   ],
   "dur_s": 13.12,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 211112,
   "md5": "a912407d647d1a6b85adc0304637fb7d"
  },
  {
   "order": 42,
   "id": "seg-042-log-centralization",
   "file": "voice/seg-042-log-centralization.mp3",
   "kind": "story",
   "category": 6,
   "chapter": "Log centralization",
   "tags": [
    "observability"
   ],
   "dur_s": 29.2,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 468157,
   "md5": "dcc872a0ca0057dfc6f1b50f80cbe163"
  },
  {
   "order": 43,
   "id": "seg-043-cloudwatch-omni",
   "file": "voice/seg-043-cloudwatch-omni.mp3",
   "kind": "story",
   "category": 6,
   "chapter": "CloudWatch Omni",
   "tags": [
    "observability"
   ],
   "dur_s": 11.6,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 186870,
   "md5": "8eb45cd5517d1748c23246e44c76a2f1"
  },
  {
   "order": 44,
   "id": "seg-044-alarm-tags",
   "file": "voice/seg-044-alarm-tags.mp3",
   "kind": "story",
   "category": 6,
   "chapter": "Alarm tags",
   "tags": [
    "observability"
   ],
   "dur_s": 20.0,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 321035,
   "md5": "0351d597800c6104b51bbda4628d0a18"
  },
  {
   "order": 45,
   "id": "seg-045-ecs-gateway",
   "file": "voice/seg-045-ecs-gateway.mp3",
   "kind": "story",
   "category": 7,
   "chapter": "ECS Gateway",
   "tags": [
    "containers"
   ],
   "dur_s": 73.04,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 1169910,
   "md5": "ebf858ea769376ec29f0d9b2d40a9cb0"
  },
  {
   "order": 46,
   "id": "seg-046-madpot-php2shell",
   "file": "voice/seg-046-madpot-php2shell.mp3",
   "kind": "story",
   "category": 7,
   "chapter": "MadPot PHP2Shell",
   "tags": [
    "security"
   ],
   "dur_s": 14.96,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 240369,
   "md5": "41a37f178752d72e702064be8f0c0b84"
  },
  {
   "order": 47,
   "id": "seg-047-s3-access-logs",
   "file": "voice/seg-047-s3-access-logs.mp3",
   "kind": "story",
   "category": 7,
   "chapter": "S3 access logs",
   "tags": [
    "storage"
   ],
   "dur_s": 20.16,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 323543,
   "md5": "5bf646524b5598b30d888ee458e1b1d7"
  },
  {
   "order": 48,
   "id": "seg-048-registration-forms",
   "file": "voice/seg-048-registration-forms.mp3",
   "kind": "story",
   "category": 8,
   "chapter": "Registration forms",
   "tags": [
    "messaging"
   ],
   "dur_s": 13.44,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 216127,
   "md5": "c4796663a624b661e015910d8771d159"
  },
  {
   "order": 49,
   "id": "seg-049-rcs-wizard",
   "file": "voice/seg-049-rcs-wizard.mp3",
   "kind": "story",
   "category": 8,
   "chapter": "RCS wizard",
   "tags": [
    "messaging"
   ],
   "dur_s": 14.08,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 226158,
   "md5": "d0a0c15cc9cab5f7beb11dfe76a58dac"
  },
  {
   "order": 50,
   "id": "seg-050-brand-profiles",
   "file": "voice/seg-050-brand-profiles.mp3",
   "kind": "story",
   "category": 8,
   "chapter": "Brand Profiles",
   "tags": [
    "messaging"
   ],
   "dur_s": 15.6,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 250818,
   "md5": "ed6f79c0dbeb2ca63f73f82bf1ad50a8"
  },
  {
   "order": 51,
   "id": "seg-051-notify-otp",
   "file": "voice/seg-051-notify-otp.mp3",
   "kind": "story",
   "category": 8,
   "chapter": "Notify OTP",
   "tags": [
    "messaging"
   ],
   "dur_s": 19.2,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 308078,
   "md5": "171d6da392763f43a5721b89644a2c9a"
  },
  {
   "order": 52,
   "id": "seg-052-azure-code-conversion",
   "file": "voice/seg-052-azure-code-conversion.mp3",
   "kind": "story",
   "category": 9,
   "chapter": "Azure code conversion",
   "tags": [
    "migration"
   ],
   "dur_s": 20.0,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 321035,
   "md5": "b8df40b4ba52031c87c79131c66499aa"
  },
  {
   "order": 53,
   "id": "seg-053-landing-zones",
   "file": "voice/seg-053-landing-zones.mp3",
   "kind": "story",
   "category": 9,
   "chapter": "Landing zones",
   "tags": [
    "migration"
   ],
   "dur_s": 16.88,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 271298,
   "md5": "e254f7c45053dc3ed4e83002cae90436"
  },
  {
   "order": 54,
   "id": "seg-054-network-migration",
   "file": "voice/seg-054-network-migration.mp3",
   "kind": "story",
   "category": 9,
   "chapter": "Network migration",
   "tags": [
    "migration"
   ],
   "dur_s": 16.24,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 260849,
   "md5": "1893ca02b0f69431a6002481f865899a"
  },
  {
   "order": 55,
   "id": "seg-055-polly-voices",
   "file": "voice/seg-055-polly-voices.mp3",
   "kind": "story",
   "category": 10,
   "chapter": "Polly voices",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 14.8,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 237861,
   "md5": "c2a018b4c4389677e62351793f929d96"
  },
  {
   "order": 56,
   "id": "seg-056-quick-singapore",
   "file": "voice/seg-056-quick-singapore.mp3",
   "kind": "story",
   "category": 10,
   "chapter": "Quick Singapore",
   "tags": [
    "ai-ml"
   ],
   "dur_s": 8.4,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 135461,
   "md5": "450780eba2dd1c81155ef35ad0e2b9df"
  },
  {
   "order": 57,
   "id": "seg-057-cloudformation-rollback",
   "file": "voice/seg-057-cloudformation-rollback.mp3",
   "kind": "story",
   "category": 10,
   "chapter": "CloudFormation rollback",
   "tags": [
    "devops"
   ],
   "dur_s": 20.72,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 332738,
   "md5": "a6b3714b1e06b99a7b167678a2202df3"
  },
  {
   "order": 58,
   "id": "seg-058-solution-matching",
   "file": "voice/seg-058-solution-matching.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "Solution Matching",
   "tags": [
    "partner"
   ],
   "dur_s": 14.08,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 226158,
   "md5": "6f57eb0964a8dd5ac143736d88422032"
  },
  {
   "order": 59,
   "id": "seg-059-program-navigator",
   "file": "voice/seg-059-program-navigator.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "Program Navigator",
   "tags": [
    "partner"
   ],
   "dur_s": 14.08,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 226158,
   "md5": "2f1b63905b064b42ec1a1d952a53db5b"
  },
  {
   "order": 60,
   "id": "seg-060-clean-rooms-databricks",
   "file": "voice/seg-060-clean-rooms-databricks.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "Clean Rooms Databricks",
   "tags": [
    "analytics"
   ],
   "dur_s": 13.52,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 217381,
   "md5": "15e7ccf20acb22c26483983ab364ccce"
  },
  {
   "order": 61,
   "id": "seg-061-global-resiliency",
   "file": "voice/seg-061-global-resiliency.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "Global Resiliency",
   "tags": [
    "contact-center"
   ],
   "dur_s": 17.76,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 285091,
   "md5": "7560d07f8ab59cbf7c31a800e82b3955"
  },
  {
   "order": 62,
   "id": "seg-062-service-reference",
   "file": "voice/seg-062-service-reference.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "Service Reference",
   "tags": [
    "security"
   ],
   "dur_s": 11.92,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 191886,
   "md5": "51908b728599faf375bbc48c47f69cb1"
  },
  {
   "order": 63,
   "id": "seg-063-mediaconnect-router",
   "file": "voice/seg-063-mediaconnect-router.mp3",
   "kind": "story",
   "category": 11,
   "chapter": "MediaConnect Router",
   "tags": [
    "media"
   ],
   "dur_s": 13.36,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 214874,
   "md5": "997de98e7cdc6b2ecc8cbece9cdd0ab6"
  },
  {
   "order": 64,
   "id": "seg-064-siliconangle-trainium",
   "file": "voice/seg-064-siliconangle-trainium.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "SiliconANGLE Trainium",
   "tags": [
    "ai-ml",
    "community"
   ],
   "dur_s": 36.32,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 582260,
   "md5": "c100e9863f44c245e104f754a3800dab"
  },
  {
   "order": 65,
   "id": "seg-065-new-stack-pytorch",
   "file": "voice/seg-065-new-stack-pytorch.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "New Stack PyTorch",
   "tags": [
    "ai-ml",
    "community"
   ],
   "dur_s": 38.24,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 612771,
   "md5": "362de1248b6f9855275a20ce2db50c0a"
  },
  {
   "order": 66,
   "id": "seg-066-repository-screening",
   "file": "voice/seg-066-repository-screening.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "Repository screening",
   "tags": [
    "security",
    "community"
   ],
   "dur_s": 49.12,
   "host": "Ken",
   "voice_name": "Liam - Energetic, Social Media Creator",
   "bytes": 787060,
   "md5": "f6c83a36459e413dad16abc8688ad368"
  },
  {
   "order": 67,
   "id": "seg-067-ai-spending",
   "file": "voice/seg-067-ai-spending.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "AI spending",
   "tags": [
    "cost",
    "community"
   ],
   "dur_s": 42.64,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 683406,
   "md5": "838cc0c131bede9d6cf9d59e8a1805b8"
  },
  {
   "order": 68,
   "id": "seg-068-netflix-identity",
   "file": "voice/seg-068-netflix-identity.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "Netflix identity",
   "tags": [
    "security",
    "community"
   ],
   "dur_s": 81.68,
   "host": "Ada",
   "voice_name": "Jessica Anne Bogart - Chatty and Friendly",
   "bytes": 1307837,
   "md5": "380389e393814fc126077049abc290b7"
  },
  {
   "order": 69,
   "id": "seg-069-security-coverage",
   "file": "voice/seg-069-security-coverage.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "Security coverage",
   "tags": [
    "security",
    "community"
   ],
   "dur_s": 14.48,
   "host": "Dennis",
   "voice_name": "Roger - Laid-Back, Casual, Resonant",
   "bytes": 232846,
   "md5": "838a48e4e949e1d5d0d7bae59c1036ac"
  },
  {
   "order": 70,
   "id": "seg-070-deployment-checks",
   "file": "voice/seg-070-deployment-checks.mp3",
   "kind": "story",
   "category": 12,
   "chapter": "Deployment checks",
   "tags": [
    "devops",
    "community"
   ],
   "dur_s": 61.12,
   "host": "Grace",
   "voice_name": "Hope - Clear, Relatable and Charismatic",
   "bytes": 978903,
   "md5": "88ac24275243b7429d6fc167909ad408"
  },
  {
   "order": 71,
   "id": "seg-071-outro",
   "file": "voice/seg-071-outro.mp3",
   "kind": "outro",
   "category": null,
   "chapter": "Outro",
   "tags": [
    "show-outro"
   ],
   "dur_s": 34.56,
   "host": "Linus",
   "voice_name": "Jarnathan - warm, confident, versatile",
   "bytes": 553838,
   "md5": "3d8973600f3a7b04fa81e42cf781052a"
  }
 ]
}
```
