Goodwin is seeking a Director of AI to lead the design and deployment of proprietary AI models to accelerate the way attorneys deliver value to clients by embedding AI into the heart of the practice, with a focus on contract analysis, M&A due diligence acceleration, litigation support, pricing forecasting, and internal operations.
Requirements
- 10+ years building ML/AI products; 4+ years leading applied ML/LLM teams
- Hands-on with modern LLM stacks, vector search, orchestration frameworks
- Strong data engineering foundations (Python, PyTorch, Spark, SQL); MLOps (Docker, feature store, model registry)
- Demonstrated experience with high sensitivity data and regulated deployments (privacy, privilege and auditability)
- Proven record translating ambiguity into shipped systems.
- Experience with CI/CD models, evaluations, observability, rollback and cost tracking
Responsibilities
- Ship proprietary models for a list of targeted use cases with documented accuracy, latency and ROI targets.
- Stand up a secure AI data layer (feature store, vector DB, retrieval pipelines) over firm content with permissions-aware retrieval.
- Develop auditable governance including model cards, evaluation harness, bias testing, and quarterly review with legal and information security.
- Operationalize CI/CD models / agents, evaluations, observability, rollback and cost tracking
- Lead end to end lifecycle including problem framing, data pipelines, model selection, fine tuning, evals, deployment, monitoring and iteration
- Architect RAG over privilege content with strict access controls. Implement safety layers including prompt hardening and output filtering
- Build evaluation and assurance reference sets from attorney labeled data, offline/online A/B and human in the loop review.
Other
- Hire and mentor a blended team to machine learning, data and prompt engineers
- Partner with KM on domain taxonomies, embeddings strategies, and feedback loops to continuously improve models.
- Work with general counsel and privacy operations on confidentiality, data retention, vendor due diligence and AI policy
- Own change management and adoption including training plans, playbooks and success metrics with named practice leaders.
- Publish internal guidance on model cards, decision logs, user guidelines.
