Research

Research. Not just wrappers.

EconLab develops independent methods for Legal-AI. Clausa is a research product — not prompting over third-party APIs.

EconLab AI Research

EconLab researches AI systems for regulated industries. The core of our work: deterministic output in domains where interpolation is dangerous — law, medicine, finance. Clausa is our first product from this research area.

Research Areas

Neuro-Symbolic Architecture for Legal Analysis: Combining LLM parsing with deterministic rule engines for German contract types.
German Legal NLP: Building and curating training and evaluation data for BGB tenancy law, KSchG, TzBfG — focused on legal precision.
Clause Classification: Automatic recognition and typing of contract clauses according to German legal categories.
Compliance Checking: Rule-based validation of clauses against court decisions and statutory texts.

External Research References

DACL Benchmark (arXiv:2601.06181): External study on the accuracy of AI systems analysing German-language court rulings. Clausa draws on this methodology for internal evaluations — the published study figures are third-party data, not Clausa's own measurements.
Legal NLP Research Landscape: Neuro-symbolic approaches for legal systems are actively researched at universities in DE/AT/CH (incl. TU Wien, LMU Munich). Clausa monitors these developments and will publish own methods incrementally.

Our Methodology

Every rule change in Clausa passes through: (1) legal expert review, (2) testing against known case decisions, (3) rejection on edge-case ambiguity. We publish quality limits before rolling out features.

Frequently Asked Questions

Does EconLab publish research results?

We publish incrementally — validated internally first, then publicly. A technical whitepaper on the Symbolic Compliance Engine is in preparation.

Does Clausa collaborate with universities?

We are in talks with Legal NLP research groups in Germany and Austria. Formal collaborations are in the pipeline.

How does EconLab differ from other Legal AI labs?

Our focus is exclusively on German-language law with deterministic output. We do not build general LLM assistants — we build vertical systems with verifiable outputs.

Receive Research Updates

Join the pilot list — we keep you informed on research progress and launch.