AI Engineer - Legal Search

Omnilex is building AI-powered legal research tools with roots at ETH Zurich. The role focuses on building and shipping retrieval, reasoning and context engineering for legal research.

Core Responsibilities:

• Implementing domain-specific retrieval and reranking algorithms beyond standard approaches
• Designing production-grade LLM systems and chatbots
• Creating scoring features from citations, authority, recency, jurisdiction, and document structure
• Managing cost/latency tradeoffs through caching and batching strategies
• Defining evaluation sets and conducting ablation studies
• Tuning search indices, analyzers, and embeddings
• Collaborating with legal experts to translate user needs into ranking features

Must-Have Requirements:

• Strong production experience with search/retrieval systems (hybrid retrieval, reranking, query understanding)
• Proven LLM product deployment from prototype to production
• Solid algorithms foundation (data structures, complexity, graph theory, statistics)
• TypeScript/Node.js proficiency
• Experience with Azure AI Search, pgvector/PostgreSQL, or similar platforms
• Understanding of embedding models and cross-encoder reranking
• Full-time availability; minimum 2 days/week on-site in Zurich
• English fluency

Nice-to-Have:

• Swiss work permit or EU/EFTA citizenship
• German language proficiency
• Evaluation pipeline experience (AI judging, human-in-the-loop labeling)
• Knowledge of sparse methods (BM25) and dense models (e5, BGE)
• Docker/Kubernetes/serverless experience
• Familiarity with Azure/NestJS/Next.js
• Legal systems knowledge

Tech Stack: AI, LLM, API, Azure, PostgreSQL, NestJS, TypeScript, Docker, SQL, Node.js, Machine Learning

Salary: CHF 100,000-140,000/year + ESOP

Benefits:

• Hybrid work arrangement (minimum 2 days on-site in Zurich)
• Remote option 1 day weekly
• Equity stake through employee stock options
• Direct impact on product quality
• Autonomy in shaping legal research pipeline
• Interdisciplinary team environment