botirk

6 ms·

1.
Show HN: GitHub Actions self-hosted runners on Modal Sandboxes(github.com)
3 points by botirk 1mo ago|1 comments
2.
by botirk 1mo ago
was looking for a long time for gpu-native scalable and highly available runners, decided to build it myself on modal. it's self hosted so you can tinker with it say what resources you need explicitly and pay for what you need only :)
3.
by botirk 8mo ago
We analyzed per-task results on SWE-Bench Verified and noticed a pattern that aggregate leaderboard scores hide: many tasks failed by the top-performing model are consistently solved by other models. For example, Claude Opus 4.5 solves the
4.
Nordlys Hypernova: 75.6% on SWE-Bench(nordlyslabs.com)
1 points by botirk 8mo ago|2 comments
5.
by botirk 8mo ago
We propose a new architecture called Mixture of Models (MoM) to solve LLM routing for coding workflows. We use a embedding + clustering approach on SWE data and then evaluate LLMs on each cluster to find out who is best.
6.
by botirk 8mo ago
cool!
7.
by botirk 1y ago
thanks!
8.
by botirk 1y ago
yea exactly!
9.
by botirk 1y ago
were you trying to build with 0.15, it was broken till recently, I like to preface this is not a feature ready OS, its just a learning experiment!
10.
by botirk 1y ago
yes good point, README was AI generate not going to lie, missed those details its updated now to work with 0.15!
11.
by botirk 1y ago
yes exactly, its based off that!
12.
by botirk 1y ago
this is not usable at all, its just to show people that OSes are not mysterious things, at a bare bones level quite simple
13.
by botirk 1y ago
no libs at all, only std where appropriate, and there is no hidden sys calls
14.
by botirk 1y ago
its been ported, currently basic paging, scheduling is implemented, but still need to implement sys call interface
15.
by botirk 1y ago
We built our infra on Azure during a hackathon. It made sense at the time, so we stuck with it. For a while, Container Apps worked fine. Then we launched our AI model router demo, and everything changed. In just two days, we spent over $250
16.
by botirk 1y ago
I hit this while building Adaptive, a proxy layer for LLM APIs. I needed to extend the OpenAI SDK / Anthropic SDK types with some extra fields. In most languages, this is trivial. In Go, it meant: → Embedding the original struct and ho
17.
Lessons from building an intelligent LLM router(github.com)
1 points by botirk 1y ago|1 comments
18.
by botirk 1y ago
We have been experimenting with routing inference across LLMs, and the path has been full of wrong turns. Our first attempt was to just use a large LLM itself to decide routing. It was too costly and the decisions were unreliable. Next we t
19.
by botirk 1y ago
I built FastKey, a Redis-compatible key-value store written from scratch in C. It started as a project to understand Redis internals. It turned into a complete implementation with real-world features. Key features: → Full RESP protocol comp
20.
by botirk 1y ago
Claude Code has exploded in popularity as a developer tool. The problem is cost, running everything directly through Anthropic gets expensive fast. We built Adaptive, a model routing platform that integrates with Claude Code as a drop-in re
21.
Show HN: Model Routing for Claude Code – same experience, just cheaper(docs.llmadaptive.uk)
1 points by botirk 1y ago|0 comments
22.
Show HN: I wrote an OS in 1000 lines of Zig(github.com)
151 points by botirk 1y ago|34 comments
23.
by botirk 1y ago
I wanted to understand what the bare minimum of an operating system looks like. So I built one in Zig, keeping the whole thing under 1000 lines of code. It can: → Boot from GRUB → Manage memory → Schedule simple tasks → Output text to VGA T
24.
by botirk 1y ago
done thanks bro!
25.
Show HN: Semanticcache – A high-performance semantic caching library for Go(github.com)
4 points by botirk 1y ago|2 comments