About the job
Why Join Omnilex?
At Omnilex, we are dedicated to revolutionizing the legal profession. Our innovative AI-driven platform empowers legal professionals to significantly boost their efficiency in legal research and automate essential workflows. We thrive on close collaboration with our clients, enabling rapid iterations that keep us at the forefront of the market. In just one year, we have transitioned from an initial MVP to a widely utilized product, benefiting thousands of legal practitioners across Switzerland, Germany, and Liechtenstein, and we are now expanding throughout Europe.
Our distinct approach tackles unique challenges by integrating external data, customer-specific insights, and our proprietary AI-focused legal commentaries.
Join our vibrant and passionate team of 15, with a strong foundation from ETH Zurich.
Your Role
Are you passionate about enhancing user search experiences? Do you excel in ranking algorithms and query understanding, and are you eager to implement improvements that users can appreciate immediately? If you're skilled at creating practical, low-latency, cost-effective solutions for AI-powered legal research, where citation accuracy, precision, and traceability are crucial, then we want to hear from you!
Key Responsibilities
As an AI Engineer specializing in Legal Search Optimization, your focus will be on developing and deploying retrieval, reasoning, and context engineering solutions that enhance our legal research capabilities.
- Implement and refine domain-specific retrieval and ranking algorithms, moving beyond traditional methods, including knowledge graphs and tailored workflows.
- Design and create robust, production-ready LLM systems and chatbots.
- Develop scoring features based on citations, authority, recency, jurisdiction, and document structure.
- Evaluate practical decisions such as API versus self-hosting and incorporate batching, early-exit, and caching mechanisms to manage costs and latency.
- Establish offline evaluation sets, conduct rapid ablations, and monitor production feedback and performance dashboards.
- Optimize search infrastructure by fine-tuning indices, analyzers, and embeddings while managing recall/precision trade-offs and eliminating duplicate entries.
- Control costs and performance by managing token usage, GPU/CPU time, and indexing expenses through caching and pre-computation strategies.
- Collaborate with legal experts to transform user challenges into actionable technology solutions.
