[bounty] support for video and voice LLM in search, timeline, meeting

[louis030195]

likely need to break down in multiple bounties

/bounty 400

eg

• meeting: use voice LLM to transcribe or summarize audio would increase a lot quality - 10x better than granola etc
• search: use video LLM would be much more powerful and different context windows
• timeline: same

suggest rough design, might create other issues

[algora-pbc[bot]]

💎 $400 bounty • screenpi.pe

Steps to solve:

1. Start working: Comment /attempt #1142 with your implementation plan
2. Submit work: Create a pull request including /claim #1142 in the PR body to claim the bounty
3. Receive payment: 100% of the bounty is received 2-5 days post-reward. Make sure you are eligible for payouts

❗ Important guidelines:

• To claim a bounty, you need to provide a short demo video of your changes in your pull request
• If anything is unclear, ask for clarification before starting as this will help avoid potential rework
• Low quality AI PRs will not receive review and will be closed
• Do not ask to be assigned unless you've contributed before

Thank you for contributing to mediar-ai/screenpipe!

Attempt	Started (UTC)	Solution	Actions
🟢 @BenraouaneSoufiane	Aug 05, 2025, 10:31:57 AM	WIP
🟢 @7908837174	Oct 23, 2025, 04:49:57 AM	WIP

[kumarvivek1752]

I wanna work on it, how are you validating this? need more context.

[RaghavArora14]

/attempt #114

[algora-pbc[bot]]

@RaghavArora14: We appreciate your enthusiasm but since you already have 3 active bounty attempts, we're going to keep this open for other contributors to attempt. 🫡

[ToSeven]

/attempt https://github.com/mediar-ai/screenpipe/pull/114

[BenraouaneSoufiane]

/attempt #1142

[BenraouaneSoufiane]

@louis030195 Proposed breakdown, would be ~400$ each:

1. Voice LLM for meetings
   • Whisper → Transcription
   • LLM summarization
2. Video LLM for search
   • Frame/audio analysis → LLM → Embeddings
   • FAISS-powered semantic search
3. Timeline enhancement
   • Combine visual/audio tags
   • Auto-label segments (topic, scene, speaker)

Would start with #1 (meeting voice summary) and propose incremental PRs. Feedback welcome.

[BenraouaneSoufiane]

@louis030195 can you release the amount?

[Deng-Xian-Sheng]

问题解决了吗？ Is the problem solved?

[kallal79]

/attempt https://github.com/mediar-ai/screenpipe/issues/1142