AI and nethermind:

[ahmohamed]

RAG system explaining ZK, high-level and technical

straighforward implementation, using together AI open-source model:

1. Collect all materials relevant to ZK
2. All source code for contracts. (we can inject full github repos)
3. Embed that in a vector DB
4. Integrate with langchain in a RAG system

AI model router:

Route API requests to different models based on permissionless transparent router. The router should incoporate feedback from user.

A few remarks below:

• What's the mechanism that will determine model selection? Obvious ones are cost and user feedback.
• A good start is to limit the scope to Openai API-compatible providers. This includes openAI, together.ai, groq, which gives us a large set of proprietary and open-source models.
• Can we integrate this with DA / Filecoin etc for secure RAG applications?
• We can deploy on Base for low fees (and bounty).
• We need to find a ZK entry into this as well.