Agentic AI Engineer — Healthcare AI

Deloitte seeks an Agentic AI Engineer — Healthcare AI in Jersey City, NJ (onsite) to design, build, and operationalize end-to-end agentic systems for healthcare decisioning across payers, providers, and life sciences.

Responsibilities

• Design and implement agentic systems with multi-step reasoning, planning, tool use, and workflow execution for complex, regulated healthcare processes.
• Build stateful workflows using LangGraph and LangChain, including branching, retries, self-correction, human-in-the-loop checkpoints, and reusable orchestration patterns.
• Engineer for long-horizon reliability: multi-step task completion, recovery from cascading errors, planning under uncertainty, and robust tool use when steps fail.
• Develop the reasoning behind regulated decisions with policy- and criteria-grounded outputs, structured proposer/critic/judge-style reviews, and auditable rationales for clinical review, prior authorization, claims integrity, and care management.
• Create end-to-end Retrieval-Augmented Generation (RAG) pipelines: ingestion, chunking, embeddings, vector/hybrid retrieval, reranking, contextual compression, and grounding strategies.
• Engineer memory and context management: conversational state, persistent memory, retrieval-aware context assembly, and token-efficient context selection.
• Apply context-delivery patterns to ensure agents access the right information at the right time.
• Implement observability and tracing for prompts, tool calls, retrieval quality, agent traces, failures, drift, latency, and production behavior.
• Apply guardrails, safety controls, and failure-handling to reduce hallucinations and unsafe actions.
• Evaluate agents at trajectory and task levels: multi-step task success, failure-mode and regression analysis, sandboxed test environments, plus retrieval- and generation-quality metrics, automated checks, and human review.
• Engineer healthcare-grade safety: deployment eval gates, human oversight and escalation models, auditability and traceability for regulated decisions, and PHI/HIPAA-aware data handling.
• Build integrations with internal and external tools, APIs, enterprise systems, databases, and model providers to operate safely within real business workflows.
• Deliver production-quality code with strong testing, CI/CD, logging, versioning, and documentation; balance quality, safety, latency, cost, and model risk in architecture decisions.
• Collaborate with modeling and post-training engineers to improve model behavior for tool use, grounding, and long-horizon reasoning through evaluation-driven feedback and, where helpful, fine-tuning or reasoning-optimized models.
• Translate ambiguous, high-complexity operational processes into robust system logic and reusable AI patterns; stay current with advances in agentic systems and translate research into practical engineering decisions.

Requirements

• Bachelor's degree in Computer Science, Engineering, Data Science, Computational Linguistics, or related field.
• Proven depth building and shipping production agentic systems; this is the primary craft, with strong software/ML fundamentals and substantial, recent hands-on agentic work.
• Hands-on experience building production agent systems with modern orchestration such as LangGraph/LangChain or equivalents, including custom orchestration.
• Experience designing and optimizing end-to-end RAG systems: indexing, retrieval, reranking, grounding, and evaluation.
• Strong memory and context management understanding: context windows, retrieval-driven context assembly, persistent memory, and high-signal context selection.
• Deep, practical knowledge of LLM behavior, including strengths, limitations, hallucination risks, reasoning constraints, and latency/cost trade-offs; plus corresponding evaluation methods.
• Experience evaluating and debugging agent behavior: task-success and trajectory analysis, not just output quality.
• Strong Python engineering skills and modern software practices: testing, CI/CD, version control, API integration; experience implementing observability, tracing, and debugging for production LLM-based systems.
• Hands-on experience with at least one frontier model platform (e.g., Anthropic, Google, OpenAI) and/or open-weight/self-hosted models (e.g., Llama via vLLM), including production tool use and agent capabilities.
• Ability to travel 0-50% on average, depending on work and client engagements.
• Limited immigration sponsorship may be available.

Technologies

• LangGraph, LangChain
• Python
• Llama via vLLM (open-weight/self-hosted models)
• Pinecone, Weaviate, Milvus
• Anthropic, Google, OpenAI
• FHIR

Preferred Qualifications

• Experience with multi-agent systems and agent collaboration patterns.
• Familiarity with vector databases and retrieval infrastructure such as Pinecone, Weaviate, or Milvus.
• Exposure to model adaptation and fine-tuning methods like LoRA or QLoRA.
• Understanding of traditional NLP concepts: tokenization, semantic similarity, entity extraction, summarization, and transformer fundamentals.
• Experience operating in regulated, high-stakes, or operationally complex environments; healthcare exposure or standards such as FHIR is a plus, not required.
• demonstrated habit of staying current with AI research, benchmarks, and emerging engineering patterns.

Compensation

• Estimated base salary range: $134,500-$265,100 (not adjusted for geographic differential). Actual base pay depends on your skills, experience, and level.