Show HN: VectorWave – Turn Python function execution into searchable vector data

[cozymori]

Hello HN,

I built an open source Python framework called VectorWave. It connects your code execution flow directly to a Vector Database (Weaviate).

Instead of dumping inputs/outputs into unstructured text logs, VectorWave embeds them as vectors. This allows you to: 1. Search execution history semantically (e.g., find "bad input" errors without exact keyword matching). 2. Implement semantic caching (skip expensive calls if similar inputs were processed). 3. Enable "Self-Healing" by comparing error traces with past successful runs.

It is designed for AI engineers needing long-term memory for agents, or backend devs debugging complex workflows.

Repo: https://github.com/cozymori/vectorwave Docs: https://cozymori.github.io/vectorwave-docs/

I'd love to hear your feedback on the concept of "Execution RAG".

[dmarwicke]

wait this seems expensive? why vectorize everything when postgres is sitting right there. is the point to search through logs or something

[cozymori]

First, we don't actually vectorize everything.

We use a hybrid approach to keep costs low and performance high:

Function Metadata: Vectorized only once (for semantic search).

Execution Logs: Stored as standard structured data without embeddings.

Since we don't embed every log, the storage cost is virtually the same as using Postgres. The real value of VectorWave is automating this 'Vector for Search