Lead (Agentic AI)

Responsibilities

• Translate business needs into workflows, choosing deterministic automation, single-agent, or multi-agent designs based on reliability, complexity, cost, and value. 
• Build modular, tested Python services that integrate large language models (LLMs), APIs, databases, and enterprise tools into production-ready applications. 
• Implement agent orchestration, including tool calling, structured outputs, state, memory, routing, retries, timeouts, and human approval steps, to ensure reliable agent behavior. 
• Build retrieval-augmented generation (RAG) pipelines covering ingestion, chunking, embeddings, reranking, and access-aware retrieval to improve knowledge grounding and accuracy. 
• Establish evaluation datasets and automated tests for task completion, accuracy, retrieval quality, tool selection, safety, latency, and cost. 
• Own deployment readiness, CI/CD, observability, versioning, and rollback, and lead incident investigation to ensure reliable production operation. 
• Implement security and governance controls, including least privilege access, PII protection, secrets management, audit trails, and prompt-injection defenses, to safeguard enterprise AI systems. 
• Mentor the team, review code and architecture, and set standards for testing, modularization, and documentation to strengthen team capability. 
• Define acceptance criteria and benchmark models and frameworks for quality, portability, and total cost to guide product and technology decisions.

Qualifications

• Educational Background: 

  • Relevant degree in Computer Science or a related field, or equivalent practical experience. 

     

• Experience: 
  • 7+ years of experience in software engineering, machine learning engineering, or applied AI, including demonstrated ownership of an LLM-powered production application from architecture through operation. 

  • 3-4 years of experience in mentoring engineers and guiding a multidisciplinary team. 

     

• Technical Proficiency: 
  • Strong Python skills, with experience in APIs, asynchronous processing, typed data models, automated testing, and Git 
  • Practical experience with agent workflows, tool/function calling, structured outputs, and retrieval-augmented generation (RAG) 
  • Understanding of LLM limitations, including hallucinations, context constraints, and nondeterministic behavior 
  • Experience with SQL, data pipelines, and vector search or other retrieval systems 
  • Experience with Docker, CI/CD, cloud deployment, monitoring, and production troubleshooting 
  • Familiarity with LangGraph, LangChain, Semantic Kernel, or equivalent agent tooling and Model Context Protocol (MCP) integrations.