Gravespeaker: Giving the Person Behind the Headstone a Voice
Post · https://dev.to/anchildress1/gravespeaker-giving-the-person-behind-the-headstone-a-voice-329e
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the...
First seen by Oriel on .
AI summary
Gravespeaker is a mobile-friendly web app that turns headstone photos into cited, AI-generated first-person stories and ElevenLabs narration while refusing unsupported biographical claims.
- Gemma 4 26B extracts names, dates, maiden names, and multiple occupants; a 20-photo benchmark scored 57/58 fields, misreading worn 1818 as 1848.
- Mastra orchestrates deterministic SerpApi searches, retries, stopping rules, and replayed saved responses, while matching code requires names plus date evidence for researched stories.
- Gemma extracts facts with source quotes; code verifies quotes, rejects contradictory pages, and falls back to snippets, inscription-only stories, or a plain epitaph.
- The 379-test unit suite covers namesakes, conflicting dates, failed searches, and unsupported claims, while a Gemini Flash evaluator distinguishes sourced life claims from theatrical atmosphere.
- ElevenLabs streams Flash v2.5 MP3 narration using Callum or Matilda with fixed speed and tunnel effects; users can switch the inferred narrator voice.
#gemma #mastra #elevenlabs #serpapi
Generated by gpt-5.6-sol. Check the original for details.
Metric trends
Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained
Reactions
Total public reactions reported by the source; distinct from points and likes.
- DEV Community: 11, latest observation ; +7 over observed period since
Comments
Comment count reported by the source.
- DEV Community: 1, latest observation ; +1 over observed period since
Aliases
| Source | External ID | URL |
|---|---|---|
| DEV Community | 4956279 | https://dev.to/anchildress1/gravespeaker-giving-the-person-behind-the-headstone-a-voice-329e Read discussion |