Lead / Principal Agentic AI Engineer
Location- Onsite (preferred - Pheonix, AZ, Johnson, RI, Dallas)
Rate- Open
This is the senior-most agentic engineer onsite the person who leads the technical relationship with an enterprise client and leads delivery of production agentic AI platforms that automate document intensive, decision-heavy workflows. You set the architecture, defend it in front of the client''s enterprise architects and security/compliance reviewers, and lead a blended onsite + offshore engineering pod to ship it.
8–10 years software engineering, with 3+ years architecting and shipping LLM / GenAI / agentic systems in production
. • Deep RAG experience: you''ve designed retrieval systems at scale — chunking and indexing strategy, embedding-model selection, hybrid/semantic retrieval, re-ranking, evaluation — and know the failure modes cold
. • MCP & connector experience: proven design of tool-calling / connector frameworks; able to define a reusable, config-driven onboarding pattern for heterogeneous internal and external sources.
• Multi-agent orchestration: expert with LangGraph / Google ADK / equivalent; you can reason about agent topology, state, retries, determinism, and where NOT to use an agent.
• Strong Python + FastAPI for production services; solid grasp of Java/Spring Boot and PostgreSQL to lead integration across a polyglot stack.
• LLM evaluation, guardrails, prompt strategy, and hallucination control as first-class engineering disciplines — not afterthoughts.
• Demonstrated technical leadership: led engineering pods, set standards, mentored seniors, and leading the client/stakeholder relationship.
• Experience delivering in regulated / security-constrained environments: audit, PII, SSO/RBAC, model governance.
• Excellent executive-level communication — you''ll be the technical voice in the room with the client.
Good to Have
• ML modeling depth: fine-tuning / adapter training, embedding-model evaluation and selection, classical ML for classification and confidence scoring, and building rigorous evaluation harnesses.
• Self-hosted / on-prem LLM serving experience (inference optimization, GPU-aware deployment, PII-safe inference).
• Document-AI / IDP depth: layout-aware extraction, OCR pipelines, footnote/table extraction from complex forms.
• OpenShift / Kubernetes architecture, OpenTelemetry, Datadog, and enterprise secrets management (HashiCorp Vault / CyberArk).
• Vector database architecture and tuning at scale (Qdrant, pgvector). • Prior onsite delivery / client-lead experience in the US.
Never miss an opportunity! Create an alert based on the job you applied for.