drguthals

3 ms·

1. 1.▲
   by drguthals 1y ago

   We built this at Tensorlake after struggling with brittle prompt-based extractors. This system lets you parse unstructured docs and extract contextual information quickly with a LangGraph Agent. I’d love thoughts from others building RAG ap
2. 2.▲
   Show HN: LangChain and Tensorlake = Better Document RAG(tensorlake.ai)

   3 points by drguthals 1y ago|1 comments
3. 3.▲
   by drguthals 1y ago

   One thing we’d love feedback on from HN folks: how are you currently orchestrating doc ingestion in AI pipelines? Did you build custom extractors, use open-source OCR, or go fully LLM? We’ve tried a bunch of approaches but I'm curious
4. 4.▲
   by drguthals 1y ago

   PDF -> Useful Information is what Tensorlake does ( https://tensorlake.ai ) Because PDFs are so dominate and yet each one has information in more than just text (tables, images, formulas, hand-writing, strike-throughs even), we
5. 5.▲
   by drguthals 1y ago

   "I know this goes against the current trend / state-of-the-art of using vision models to basically “see” the PDF like a human and “read” the text, but it would be really nice to be able to actually understand what a PDF file conta