CMLewis

6 ms·

1. 1.▲
   by CMLewis 6mo ago

   I see what you're saying, we may tweak the hero to make it a bit clearer. Thanks for the note!
2. 2.▲
   by CMLewis 6mo ago

   Thanks for trying it out! Yeah good catch on the demo. If this were a production deployment, the citations would be hyperlinked to object storage. Captain is just the index, so the real files would be wherever they were indexed from.
3. 3.▲
   by CMLewis 6mo ago

   You're right that you can chat with files using Gemini or a codegen'd RAG pipeline, and that does work well for a lot of teams. The problem that Captain really addresses comes when production pipelines need to run continuously ove
4. 4.▲
   by CMLewis 6mo ago

   OpenSearch provides general search infrastructure and they recently added vector search. It's a low level engine so users would still need to build their own ingestion, parsing, chunking, embeddings, re-ranking, permissions, etc. Onyx,
5. 5.▲
   by CMLewis 6mo ago

   Thanks, just shipped a fix ;)
6. 6.▲
   by CMLewis 6mo ago

   Yeah QMD is quite impressive! The main difference between us and them is the scale folks would be looking at indexing. The serverless ingestion engine I described in the post is optimized for processing large batch jobs with high concurrenc
7. 7.▲
   by CMLewis 6mo ago

   No way, that's awesome!
8. 8.▲
   by CMLewis 6mo ago

   Thanks! The largest alternative to Captain is folks trying to build file search themselves. As mentioned in the post, it is a lot to manage. The most similar product I've seen is Vertex File Search. They're hosted inside of GCP wh
9. 9.▲
   by CMLewis 6mo ago

   Great question, we have deterministic page # citations for PDF results and exact bounding box citations coming very soon. If you want to check out the Query API response example, here's a link: https://docs.runcaptain.com&#x
10. 10.▲
    by CMLewis 6mo ago

    Thanks, we're just starting to optimize more for the semi-structured data. So far, we've been parsing tables into Markdown and running them through the contextualized embedding model with no overlap, taking advantage of how it str
11. 11.▲
    Launch HN: Captain (YC W26) – Automated RAG for Files(runcaptain.com)

    57 points by CMLewis 6mo ago|38 comments