AI Workflow Architect

DeepJudge is a fast-growing venture-backed startup at the forefront of applying Artificial Intelligence in the legal domain. Founded by ex-Googlers and ETH PhDs, our mission is to help legal teams harness their collective knowledge so they can focus on the strategic aspects of their work. DeepJudge’s state-of-the-art knowledge retrieval system directs legal professionals to the internal insights that matter most, providing a strong foundation for AI-powered workflows. We’re headquartered in Zurich, with a team in the United States and Canada, and cater to clients all around the world.We’re a team of professionals who are passionate about their craft with a terrific record of applying technology in research and industry. Apart from that, we are pet lovers, cave researchers, yogis, divers, gamers, and so much moreWe're looking for an AI Workflow Architect who wants to become a part of a world-class AI / Legal Tech startup and join us in solving the biggest challenges in the legal industry, by leveraging generative AI on top of DeepJudges’s SOTA information retrieval technology.This key role bridges a deep understanding of our product with a nuanced understanding of our customers and their business processes, ensuring that your work drives value creation from day one. The ideal candidate will possess a high degree of curiosity and demonstrate a strong interest in generative AI and how it can be applied to shape the legal profession.Your responsibilities will include:Identifying opportunities and translating them into value-adding Workflows using our proprietary Workflow Builder, while advising customers to build solutions for high-value use casesEvaluating Workflow performance, refining them based on feedback, and creating benchmarks to better understand their value and impactIncorporating new techniques and functionality into existing Workflows while collaborating with engineering and ML teams to enhance our Workflow BuilderServing as a knowledge resource for stakeholders, including creating documentation, guides, and customer-facing materialsMonitoring industry developments and evaluating new models and techniques to be deployed in productionOptimizing the balance between quality and speed through prompt engineering, model selection, and architecture decisions across various deployment environmentsWhat we are looking for:Bachelor's degree in Computer Science, Business/Law, or related field with proficiency in Python (JavaScript and templating languages are a plus)Experience with frameworks like DSPy, LangChain, LangGraph or demonstrated success building with language modelsCustomer-centric mindset with exceptional curiosity, fast learning ability, and interest in developing deeper legal domain knowledgePersistent problem-solver who rises to challenges and navigates constraints effectivelyExceptional written and verbal communication skills with ability to translate technical concepts for diverse audiencesEnthusiasm for GenAI advancements and ability to evaluate their potential applications to our workflowsTeam player who thrives in fast-paced, collaborative environmentsWhat you can look forward to:Taking full ownership of your projects from day one and delivering tangible results that solve key customer challengesBeing part of an ambitious team pushing the frontiers of what's possible with GenAI in professional environmentsHaving a stake in shaping the future of a high-impact, high-growth deep-tech startupCompetitive compensation with opportunities for professional growth and career advancementA supportive and collaborative company culture that values your contributionsIf this role ignites a spark in you, but you're concerned about not meeting all requirements, we encourage you to apply anyway - At DeepJudge, we believe that smart minds learn quickly.This is important to us: DeepJudge is an equal opportunity employer, and we value diversity. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, or disability status. #J-18808-Ljbffr