We are bulding an AI-native, HIPAA- and HITRUST-compliant Revenue Cycle Intelligence platform designed to fundamentally modernize how healthcare organizations manage claims, payments, and revenue recovery. Built as a modular, agentic system, the platform ingests data from EHRs, practice management systems, through a secure integration gateway, normalizes it into a canonical schema, and then orchestrates a sequence of specialized intelligence agents across intake, eligibility, documentation understanding, coding, claim quality, submission, payments, denials, and appeals. Each agent is deterministic where compliance demands it, LLM-assisted where human-level reasoning creates value, and fully auditable through an enterprise audit trail and human-in-the-loop framework. The platform proactively prevents denials, optimizes charge and claim accuracy, automates high-confidence workflows, and elevates complex decisions, such as appeals and payer disputes, into explainable, branded, and learning-driven processes, delivering faster cash flow, higher recovery rates, and a dramatically lower operational burden for providers. ABOUT THIS ROLE The AI/ML Engineer owns the intelligence layer of the platform, a production LangGraph orchestration system that runs 20+ specialized agents across the entire claim lifecycle: eligibility verification, intelligent coding, claim scrubbing, denial analysis, appeals generation, and revenue pattern intelligence. These agents are not demos or prototypes. They process real claims, interact with real payers, and directly affect whether providers get paid. You will work on top of an architecture built by the Platform Architect, implementing agents that are deterministic where compliance demands it, LLM-assisted where human-level reasoning creates value, and fully auditable at every step. This is one of the most technically interesting AI engineering problems in healthcare; combining structured payer rules with LLM-driven document understanding and agentic workflow orchestration. WHAT WE'RE LOOKING FOR • 4+ years of production ML or AI engineering experience • Hands-on LangGraph production experience, not just familiarity, you have built and shipped agentic workflows • Strong LangChain experience — tool use, memory, retrieval-augmented generation (RAG), multi-agent orchestration • LLM prompt engineering depth — you understand how to get reliable, structured, auditable outputs from LLMs in high-stakes environments • Python expertise — you write clean, testable, production-grade Python • Experience with vector databases (Pinecone, Weaviate, pgvector) and embedding strategies • Strong understanding of ML evaluation — you know how to measure whether an agent is actually working, not just whether it runs • Ability to reason about structured + unstructured data — claims are structured, clinical notes are not, and agents need to work with both NICE TO HAVE • Healthcare domain knowledge — familiarity with ICD-10, CPT, HCPCS, payer adjudication logic, or denial management • Experience with medical coding or clinical documentation improvement (CDI) • Background in regulated AI; you understand what auditability and explainability actually require in practice • Fine-tuning experience on domain-specific models WHY THIS ROLE You will build the AI system that makes the platform fundamentally different from every legacy RCM platform on the market. You will work on genuinely hard problems at the intersection of LLM reasoning, structured healthcare data, and compliance-grade auditability. This is not a wrapper around ChatGPT. This is a serious AI engineering challenge with real clinical and financial consequences.