Applied AI Engineer - OCR, RAG/LLM, Evaluation/QA
Tech Stack
Agent Workflow
Building RAG/LLM pipelines for legal document summaries, client email drafts, and action items. Confidence scoring with rule/LLM checks and human-in-the-loop gates. OCR/extraction pipeline with evaluation framework.
About the Role
Applied AI Engineer – OCR, Document Classification, Extraction, RAG/LLM, Evaluation/QA
Company: MTR Legal Rechtsanwälte (economic law firm, mid-sized, international)
Location: 100% Remote (CET/Europe ideal)
Employment Type: Freelance | Full-time or Part-time
Start: Immediate (asap)
Published: February 2, 2026
About MTR Legal:
MTR Legal is a mid-sized, internationally-focused law firm specializing in family law. They seek to combine legal clarity with entrepreneurial dynamism while building innovative AI solutions for legal document processing.
Key Requirements:
- Production-grade LLM/RAG experience (not demos)
- Document-focused expertise (OCR/extraction/classification)
- Engineering rigor: versioning, reproducibility, testing, monitoring
- Evaluation mindset: measurable quality, regression prevention
Deliverables:
- DocTypes + extraction schemas (top 10 document types) with validations
- OCR/extraction pipeline (handling poor scans, tables, stamps)
- RAG/LLM: summaries, client email drafts, action items
- Confidence scoring: thresholds, rule/LLM-checks, human-in-the-loop gates
- QA/evaluation: golden sets, precision/recall metrics, regression tests
Tech Stack: OCR/extraction tools, LLM/RAG, embeddings, prompting, evaluation harness, monitoring/drift detection, workflow integration
Workload by arrangement. Project-based.