Senior AI Developer

Kilpatrick Townsend & Stockton LLP•Menlo Park, CA

8d•$180,000 - $250,000•Hybrid

About The Position

Kilpatrick Townsend & Stockton LLP is a premier Am Law 100 multinational law firm with 22 offices worldwide, known for its intellectual property, corporate litigation, and technology practices. The firm's AI Innovation team builds and ships production AI systems that serve hundreds of users across the firm and is expanding to a client-facing portal. We work with the latest platforms from Anthropic and Microsoft, run agentic workflows in production, and maintain a growing network of Model Context Protocol (MCP) integrations that connect AI agents to enterprise systems. This is a small team with startup-like autonomy inside a global enterprise. At Kilpatrick we are one team where each person plays an integral role in serving the needs of our clients. The firm has a strong dedication to its employees, values, and commitment to the community. We are looking for a Senior AI Developer, fully remote, to help shape the future of AI at the firm. This is an individual contributor role with significant technical ownership. You will design and deploy agentic AI systems, scale our internal platforms, extend our MCP infrastructure, and help define the technical roadmap for AI adoption. Our AI systems are in production today, serving the firm at scale. We are looking for someone who is ready to contribute immediately to an active, fast-moving development environment. Your first 6 to 12 months will focus on: Scaling document intelligence and RAG pipelines to handle millions of enterprise documents with high retrieval accuracy Expanding agentic workflows and AI skills that automate complex, multi-step legal and operational tasks Strengthening the MCP integration layer with new tool connections, improved security, and increased reliability

Requirements

Bachelor’s degree in Computer Science, Engineering, Data Science, or a related field preferred
Demonstrated experience building agentic AI systems with tool use, function calling, and multi-step workflow execution in production
Strong working knowledge of the Anthropic platform (Claude models, Messages API, streaming, extended thinking, Claude Agent SDK)
Experience with Microsoft AI services (Azure OpenAI, Azure AI AI Foundry, Semantic Kernel, or related offerings)
Experience with MCP or similar AI-to-tool integration frameworks (LangChain, LlamaIndex, CrewAI, custom pipelines, or equivalent)
Experience building RAG pipelines, semantic search, and document intelligence workflows at scale
Proficiency with real-time streaming architectures (SSE, WebSockets, Streamable HTTP) and enterprise auth (OAuth 2.0/2.1, Azure AD/Entra ID)
Ability to work across the stack: backend APIs, streaming infrastructure, frontend interfaces
Proficiency with AI-native development platforms (Claude Code, GitHub Copilot, Cursor, or similar) over expertise in any single language
Track record of independently solving complex, ambiguous problems and shipping working solutions

Nice To Haves

Understanding of ML fundamentals: fine-tuning (LoRA, QLoRA, PEFT), evaluation metrics, NLP tasks, and MLOps/LLMOps practices
Familiarity with ML frameworks (PyTorch, Hugging Face Transformers, scikit-learn)
Experience with large-scale data processing: millions of documents, vector embedding pipelines, enterprise databases
Knowledge of large law firm operations, IP workflows, or enterprise document management systems (iManage, NetDocuments)

Responsibilities

Help define the technical roadmap for AI adoption, contributing to build-vs-buy decisions, vendor evaluations, and architectural direction
Architect multi-agent systems and agentic workflows using tool use, function calling, planner-executor pipelines, agent swarms, and autonomous task decomposition
Build and maintain AI skills: self-contained automation workflows that run in their own execution context with quality assurance verification
Design context management strategies including RAG, semantic search, vector embedding pipelines, conversation compaction, long-term memory management, and grounding
Extend our MCP server network with new tool integrations, dynamic server discovery, gateway architectures, OAuth 2.1 auth flows, audit logging, and sandboxed execution
Build real-time streaming interfaces that surface AI-generated insights, citations, and structured outputs
Engineer solutions for large-scale data challenges across millions of documents, patent databases, and enterprise systems
Implement guardrails, evaluation pipelines, and data governance standards appropriate for a regulated legal environment
Evaluate and integrate new models, APIs, and capabilities as they ship from Anthropic, Microsoft, and other providers