Radar / AI infrastructure / XERJ
XERJ — local AI search engine for coding agents
Apache-2.0 local search engine (single static binary, Elasticsearch-compatible) that auto-indexes any folder — code, docs, logs, PDFs — so coding agents retrieve exact passages instead of burning context on grep.
Why it matters
Measured, not marketing: a controlled study (8 tasks, 4 languages, real billed token counts) showed 2.7x fewer output tokens than a grep-driven agent at the same 16/16 solve rate. Its 'reference coding' loop — clone the OSS repos closest to a task, index them, retrieve how they solved it — is gated honestly to code the model hasn't memorized.
What you could build with it
A dev-tools team could ship XERJ as the search backend for its internal coding-agent fleet: auto-index every private monorepo once, and let agents query passages by meaning instead of each agent re-reading whole files — cutting token spend exactly where models have no memorized knowledge.
Does it hold up?
Promising but vendor-measured — the headline 2.7x/26x token cuts come from the project's own study; the honest limits (gated by memorization) are documented in-repo. No fully independent third-party benchmark verified yet.
Built with XERJ
- XERJ reference-coding case studyxerj.org · The project's own controlled study with real billed tokens: 2.7x fewer output tokens than grep at the same solve rate, 26x fewer than memory alone, and 9/9 vs 0/9 on a Rust library the model had never seen.
- ZeyadOsama567/xerj (community fork)GitHub · Active community fork whose SKILL.md documents the honest scope — decisive win on unmemorised code, pure overhead on popular public crates — showing real downstream engagement with the benchmark.
- zeyadosama567/xerjgithub · Community fork reproducing the benchmark discussion and documenting honest scope — the main downstream fork found with real activity.
Learn more
First spotted on github: source.
More AI infrastructure
DeepGEMM: DeepSeek's open GPU BLAS kernel libraryDeepSeek's DeepGEMM, a clean and efficient open-source BLAS kernel library for NVIDIA GPUs (up to 1550 TFLOPS on…infra · JEV 0.77OpenViking — self-evolving context database for AI agentsByteDance Volcengine's open-source (AGPL-3.0) context database that unifies agent memory, knowledge RAG, and skills…infra · JEV 0.76Modular open-sources MAX inference server, Mojo stdlib and accelerator kernelsThe core of Modular's unified AI deployment stack is now open: the MAX inference server (OpenAI-compatible endpoints)…infra · JEV 0.76OmniRouteFree MIT AI gateway: one endpoint, 290+ providers and 500+ models with quota-aware auto-fallback.infra · JEV 0.75Context Mode: context-window optimization for AI coding agentsMCP server for AI coding agents that sandboxes verbose tool output and persists session memory, cutting context use by…infra · JEV 0.74StrataOpen-source inference engine that runs the 125B-parameter Qwen3.8-Flash-Next MoE on a single consumer GPU with 12GB+…infra · JEV 0.74
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.