frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Build queryable packs for AI agents from videos, podcasts, and files

https://github.com/buralog/beyin
1•buralog•1h ago
Hi,

This started from a pretty personal use case.

There was this very technical person I follow who would go live on YouTube from time to time. He has a ton of experience, and would casually drop really good insights about software architecture, engineering tradeoffs, and just general "you only learn this after years" kind of stuff. He also posts shorter clips, but I wanted something else: I wanted that knowledge to be always there, queryable whenever I needed it.

At the same time, I was also trying to understand what RAG actually is in practice, and how to learn applied AI by building something real instead of just reading about it.

My first thought was: ok, this probably has to be fully local. I assumed if I want to query my own stuff locally, then I need to use a local LLM. So I looked into Ollama and thought, alright, I can build this on top of that and just query everything on my machine. At that point I also had some pretty wrong assumptions about local models and resource usage.

After building the first version, it worked, but the result felt a bit underwhelming. Retrieval itself was useful, but the final answer didnt feel as smart as I expected. I use Codex and Claude Code a lot in my daily workflow, so maybe I was unfairly expecting something that felt more "intelligent", or at least looked that way.

Then after a lot of tests with agents (Codex and Claude Code, I use both), I realized something kind of obvious: the Ollama part was mostly just taking the retrieved chunks and turning them into a proper answer. And if that's the job, why couldnt an agent do the same thing?

So I tried wiring it through MCP.

That was the moment where the project really clicked for me. The answers became way better structured, the whole thing felt much smarter, and more importantly, it fit directly into how I already work. Instead of having a seperate tool where I go ask questions, the knowledge just becomes available inside the agent workflow itself. The agent can retrieve it, use it, suggest things, and continue the task.

That was exactly what I wanted, maybe even better than what I had in mind when I started.

The best part for me is that once it's set up, it kinda disappears. I just keep adding YouTube videos, podcasts, and files, and then that context is available while I'm working with AI agents. It stops feeling like "a RAG demo" and starts feeling like part of the actual workflow.

What started as a small local RAG experiment ended up turning into something much more useful than I originally imagined.

Ask HN: Are You Using Finetuning?

1•nate•31s ago•0 comments

ShadowStrike Phantom: Open-Source EDR/XDR Platform

1•Soocile•2m ago•0 comments

Trump is 'calling for a nuclear strike,' former White House comms director says

https://www.irishstar.com/news/politics/breaking-trumps-former-comms-director-36980303
2•ndsipa_pomu•2m ago•0 comments

Show HN: Marimo pair – reactive Python notebooks as environments for agents

https://github.com/marimo-team/marimo-pair
3•manzt•2m ago•0 comments

Making of Words.zip (Infinite Word Search)

https://luke.zip/posts/making-of-words/
1•yathern•2m ago•0 comments

Agentic development aspirations: build, run, observe – without more Markdown

https://devblogs.microsoft.com/aspire/agentic-dev-aspirations/
1•vyrotek•3m ago•0 comments

Say no to a 'camera on your face', says Meta smart glasses rival

https://www.ft.com/content/30390769-2dc1-4573-8d2e-2e8359b2ee39
1•bookofjoe•4m ago•1 comments

Got Bored due to endless scrolling on ChatGPT and Gemini

https://www.trydraft.app/
1•0xNishchit•5m ago•0 comments

What we learned about TEE security from auditing WhatsApp's Private Inference

https://blog.trailofbits.com/2026/04/07/what-we-learned-about-tee-security-from-auditing-whatsapp...
1•wslh•10m ago•0 comments

My Attempt on AI Workflow

https://github.com/jialic/pilot
1•jialic•10m ago•0 comments

This Spillway Failed on Purpose [video]

https://www.youtube.com/watch?v=UF63eFJmbrQ
3•Wilsoniumite•10m ago•0 comments

Show HN: Maintenance OS – AI-powered property maintenance for landlords

https://ifbids.com
1•arishec•10m ago•0 comments

BQN: Primitive Overloading

https://mlochbaum.github.io/BQN/commentary/overload.html
1•tosh•11m ago•0 comments

Open-Source Cannabis Price Index – Methodology, SQL, and Sample Data

https://github.com/TheoV823/cannabis-price-index
1•Tval•11m ago•0 comments

The Fermenter's Guide to Launching a Product

https://vibeagentmaking.com/blog/fermenters-guide-to-launching-a-product/
1•vibeagentmaking•11m ago•0 comments

The Building Block Economy

https://mitchellh.com/writing/building-block-economy
1•cdrnsf•14m ago•0 comments

Show HN: A reasoning hierarchical robotics pipeline you can run in the browser

https://avikde.github.io/vla-pipeline/
2•avikde•14m ago•0 comments

How good is Opus 4.6 at vuln detection?

https://zeropath.com/blog/benchmarking-opus-4-6-vuln-detection
1•NonStopOyster•15m ago•0 comments

The Building Block Economy

https://twitter.com/mitchellh/status/2041566958681014418
1•tosh•15m ago•0 comments

OpenNOW: An open-source desktop client for GeForce NOW

https://github.com/OpenCloudGaming/OpenNOW
1•takoid•15m ago•0 comments

Six (and a half) intuitions for KL divergence

https://www.perfectlynormal.co.uk/blog-kl-divergence
1•jxmorris12•15m ago•0 comments

Security versus Interoperability: Real Tension or False Dichotomy?

https://www.lawfaremedia.org/article/security-versus-interoperability--real-tension-or-false-dich...
1•hn_acker•17m ago•0 comments

Vibe Coding Tools Are a BattleMech

https://alexmeub.com/vibe-coding-tools-are-a-battlemech/
1•meub•17m ago•0 comments

A whole civilization might die tonight

https://www.nbcnews.com/politics/white-house/trump-threat-whole-civilization-will-die-iran-war-de...
17•hedayet•18m ago•8 comments

Ask HN: Why does it look like everyone is abandoning GitHub Copilot?

2•fabev•19m ago•2 comments

Get more done with new vertical tabs and immersive reading mode in Chrome

https://blog.google/products-and-platforms/products/chrome/new-chrome-productivity-features/
1•xnx•19m ago•1 comments

Show HN: Ollama-client-rs, a Rust client for Ollama

https://github.com/anperrone/ollama-client-rs
1•fushji•19m ago•0 comments

Hallucinated citations are polluting the scientific literature

https://www.nature.com/articles/d41586-026-00969-z
2•BeetleB•19m ago•0 comments

In Vivo Car T Causes Serious Toxicities in All Patients of Early Trial

https://www.insideprecisionmedicine.com/topics/oncology/in-vivo-car-t-causes-serious-toxicities-i...
2•randycupertino•24m ago•1 comments

Russia Hacked Routers to Steal Microsoft Office Tokens

https://krebsonsecurity.com/2026/04/russia-hacked-routers-to-steal-microsoft-office-tokens/
1•speckx•25m ago•0 comments