Label Sleuth
IBM's open-source no-code system for text annotation and classifier building.
About
Label Sleuth is IBM Research's open-source system for labeling text and building custom classifiers without writing code. It targets domain experts, doctors, lawyers, researchers, who need NLP models for their own data but have no ML training. As you label examples in the browser UI, models train in the background, suggest which examples to label next, and surface predictions you can correct. The EMNLP 2022 demo paper framed it as the first text-classification platform designed for non-ML audiences. On GitHub it holds a few hundred stars with research-paced maintenance; it is a research project, not a commercial product, so expect rough edges rather than polished support.
What it can do
- No-code text annotation UI built for domain experts
- Background model training while you label
- Active-learning suggestions for which examples to label next
- Custom text classifier built in hours, not weeks
- Pluggable models and components
- Open-source Apache-2.0 codebase
Setup
Clone the repo and launch with Docker Compose, or pip install label-sleuth.
Prerequisites
Docker or Python 3.9-3.11. No ML expertise required.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: 275 GitHub stars and 40 forks; IBM Research backing; EMNLP 2022 demo paper; last pushed May 2025. Strengths: genuinely open code, a clear no-code value proposition for domain experts, active-learning loop built in. Docked for: (1) modest adoption for an IBM-backed project; (2) research-paced maintenance with no commercial support. License: Apache-2.0.
★ 275 GitHub stars
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 4.0/10
- Maintenance · recency, releases, issue load
- 6.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · IBM Research open-source no-code system for text annotation and classifier building; EMNLP 2022 demo paper.
- · 275 GitHub stars, 40 forks; last pushed May 2025; research-paced maintenance.
- · Manual review: entry text written from GitHub README and IBM developer page; speedup framing kept modest, no inflated claims.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.