Vector Legal
Founding AI Engineer
New
120000 –180000 USD / yearUSAMiddlellmlangchainocrdata pipelinesevaluation frameworksprompt design
Design, build, and deploy production-grade LLM systems to power a legal platform focusing on applied AI for legal workflows.
Responsibilities
- We’re hiring an AI Engineer to design, build, and deploy production-grade LLM systems that power our legal platform. This role is focused on applied AI, turning real-world legal workflows into reliable, agent-driven systems that operate over documents, messages, and structured data.
- You’ll work at the core of our AI layer, building agents, evaluation systems, and data pipelines that directly impact how legal work is performed. The environment is highly iterative and performance-driven, with a strong emphasis on reliability, correctness, and real-world usability, not just demos.
Requirements
- 3–7 years of engineering experience, or a strong track record building and shipping AI systems (excluding internships)
- Proficiency in TypeScript and experience building production backend systems
- Hands-on experience with LLM frameworks (e.g., LangChain) and agent architectures
- Experience designing and running evals (benchmarks, test sets, regression tracking)
- Familiarity with document processing pipelines, including OCR and unstructured data extraction
- Strong understanding of prompt engineering, retrieval (RAG), and tool-augmented generation
- Ability to operate in ambiguous, fast-moving environments and own systems end-to-end
- Startup or high-velocity experience preferred
Other
- Design and deploy LLM-powered systems for document understanding, drafting, and workflow automation
- Build and orchestrate agent-based systems (e.g., LangChain / Deep Agents) for complex, multi-step tasks
- Develop and maintain evaluation frameworks to measure accuracy, reliability, and regressions
- Implement data pipelines for ingesting and structuring documents (including OCR workflows)
- Improve system performance through prompt design, retrieval strategies, and tool integration
- Work closely with product and engineering to translate legal workflows into AI-native systems
- Debug and optimize long-running agent jobs and failure modes in production
- Contribute to overall AI architecture, including memory, context management, and retrieval layers