Launching a new home-grown embedding LLM for RAG
Hi HN! Vectara is a "batteries included" retrieval augmented generation platform. You can upload your rich text documents like PDFs, HTML pages, word docs, etc, or semi-structured JSON and Vectara handles the text and metadata extraction, segmentation, vector embedding, and vector storage, and keyword storage. You can ask a question or perform a search in the UI or via our APIs and Vectara will automatically handle the vectorization, structured metadata filtering, vector+keyword retrieval, hybrid blending, and generative summarization of the results. We're focusing on building and…
In plain words
Vectara is a retrieval augmented generation platform that handles document processing, vector embedding, and storage for users who want to build AI applications. Users upload documents like PDFs and Word files, and the platform automatically extracts text, segments content, creates embeddings, and stores them. Through APIs or a web interface, users can search documents and retrieve summarized answers. The platform combines vector and keyword search methods to improve retrieval accuracy.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN! Vectara is a "batteries included" retrieval augmented generation platform. You can upload your rich text documents like PDFs, HTML pages, word docs, etc, or semi-structured JSON and Vectara handles the text and metadata extraction, segmentation, vector embedding, and vector storage, and keyword storage. You can ask a question or perform a search in the UI or via our APIs and Vectara will automatically handle the vectorization, structured metadata filtering, vector+keyword retrieval, hybrid blending, and generative summarization of the results. We're focusing on building and operationalizing the complex infrastructure for vector storage, hybrid retrieval, and generative summarization so you can use fairly high-level APIs and focus on building your own applications. We know that retrieval accuracy is incredibly important for RAG: garbage in, garbage out. We've seen a lot of projects not spend enough time on really getting the retrieval model right and wasting a lot of time/money with poor outcomes. We've spent about the past 6 months working on a new embedding model named Boomerang and just released it on the Vectara platform. We've run it through standard evaluations like BEIR (though we know many models over-fit against BEIR) as well as multi-domain evaluations. We've published the details of our tests for those that really want to dive in, but the TL;DR is that Boomerang beats most/all publicly available models in many/most situations and is particularly strong at cross-lingual and multi-domain tests. We'd love any and all feedback!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
