Matching · Validation · Review

Prove what your screening actually catches.

Nocterra Labs is building the next generation of entity matching systems - open source, AI native, and covering the full lifecycle at any scale.

Explore the tools Follow on GitHub

Console

The workbench Agent ready

A proving ground for screening engines. Load a test set, compare across engines, and measure what actually comes back.

LOAD

Adversarial by design

Curated test packs built to evade - so results reflect what an engine really catches.

RUN

Any engine, one contract

Vendor-neutral and agent ready - every workflow is an API that humans, pipelines, or AI agents can drive.

PROVE

Evidence, not vibes

Recall, false positives, latency - every run reproducible, bound into evidence your auditor can hold.

Lumen

Reference engine, built on Lucene

An open-source entity matching engine - the transparent baseline everything else is measured against.

Spec-native. Implements the screening scan spec end to end - the reference other engines are checked against.
Transparent matching. Every stage of the pipeline is inspectable. No black box.
Free, forever. The open baseline - free to run, embed, and build on.
lumen - scan API $ docker run -p 8080:8080 nocterra/lumen lumen ready on :8080 $ curl -s :8080/scan -w '%{time_total}s' \ -d '{"name":"Alexander Petrov"}' {"matches":[{"entity":"PETROV, Alexander","score":99.99}]} 0.006s