To build LLM apps on company data, retrieval augmented generation (RAG) is essential. With Linq API, you can build RAG not just for textual content, but also for images and tables - in an hour!
Thank you Abhishek for your review -- We are always looking for ways to make Linq better. Please feel free to share your thoughts with us, and stay tuned for our future updates! 😀
👋 Hello, ProductHunt community!
I’m Jacob, the Founder & CEO of Linq. We’re excited to share Linq’s one-stop-shop RAG solution, which used to be a bespoke vector search service for enterprises like Samsung Insurance but now open to the general public through a single API.
We’ve been serving domain-specific fields such as legal, insurance, and healthcare, understanding that the core of successful B2B LLM apps lies in their accuracy and relevance. All this power is packaged into a single API to be enjoyed by the wider developer community. We hope this could offer a more accessibility to advanced RAG technology.
‘Linq API for RAG’ offers:
✅Instant Implementation: A single API covering all RAG steps from extraction to re-ranking.
✅High Search Accuracy: Get accurate search results with cited sources, thanks to our augmented vector embedding and hybrid search.
✅User-friendly Knowledge Management: Easily manage and edit your knowledge base.
✅Special Multimodal AI Search Capabilities: Our search isn’t limited to texts; it also encompasses images like photos, tables, and screenshots.
How To Use ‘Linq API for RAG’:
1️⃣Upload: Upload documents or connect to your existing apps.
2️⃣Search: Search answers to your queries with cited sources.
3️⃣Generate: Generate texts and freely edit the details.
A quick heads up, Linq API is currently in its early exploration phase of deciding which features to build first - so we’re starting with PDFs for now among the many more integrations to come to help you get the feels of how it works.
We are offering a free trial until the end of 2023! Feel free to try it out! Looking forward to your candid feedback🙂
@ella_sullivan Thanks for your support! It's great to see your interest for the multimodal AI search. You're right, there's a lot in documents that isn't text, and we're tackling that.
Excited to share more with you soon – stay tuned!
