Polarityinc

Polarityinc / zenith

Public

Zenith — AI-agent-trace database in Rust

zenith.polarity.so
41
2
100% credibility
Found May 13, 2026 at 41 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

ZenithDB is an open-source columnar database engine designed specifically for efficiently storing and querying traces from AI agents.

How It Works

1
🔍 Discover ZenithDB

You hear about a simple way to store and search your AI agent's activity logs, perfect for tracking what your smart assistants are doing.

2
📥 Install with one command

Download and set up the tool on your Mac or Linux computer in seconds, no complicated setup needed.

3
🚀 Start your local database

Launch a personal database right on your machine to begin capturing traces without any extra services.

4
📤 Send your first traces

Push sample activity from your AI agents into the database and watch them arrive safely.

5
🔍 Ask questions with simple searches

Run easy questions like 'show recent errors by model' and get instant answers from your traces.

6
📊 View everything in a dashboard

Open a web page to see live stats, recent data, and run more searches visually.

✅ Monitor AI agents effortlessly

Now you have fast, reliable tracking for your AI projects, scaling from your laptop to full production.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 41 to 41 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is zenith?

Zenith is a Rust-built columnar database tailored for AI agent traces—those verbose, sparse JSON logs from agent runs packed with high-cardinality fields like prompts and models. It ingests bursts via HTTP, gRPC, or OTLP endpoints and queries them fast using SQL or ZenithQL, solving the bloat of generic observability tools on ai-agent-trace workloads. Curl the install script for a local server in seconds, data on your FS, plus a web console for live inspection.

Why is it gaining traction?

It sidesteps 10-100x inefficiencies of span-focused backends by optimizing for trace locality and late materialization, delivering sub-second queries on millions of rows. Devs love the no-Docker quickstart, curl-based ingest/query, and benchmarks beating alternatives on GitHub zenith searches for ai-agent-trace database needs. The alpha badge and Rust perf hook tinkerers building agent observability.

Who should use this?

AI teams tracing LLM/tool chains, needing cheap storage for JSON-heavy spans without full OLAP overhead. Suits prototyping agent dashboards or analytics on models, errors, and latencies via simple SQL over OTLP feeds. Skip if you're not deep in ai-agent-trace; it's niche over broad logs.

Verdict

Promising alpha for ai-agent-trace database users—strong quickstart, docs, and benches—but 41 stars and 1.0% credibility score signal early days; track CHANGELOG for stability. Try locally if agent observability bites; hold for prod otherwise.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.