Senior AI/ML Engineer- Healthcare AI platform

About RapidClaims RapidClaims is a leader in AI-driven revenue cycle management, transforming how US healthcare providers run mid- and end-revenue cycle operations from medical coding and charge capture through claim scrubbing, denials management, appeals, and payment posting. The company has raised $11 million in total funding from top investors, including Accel and Together Fund. Join us as we scale a cloud-native platform that runs self-hosted, fine-tuned Large Language Models, knowledge graphs, and embedding-based retrieval over millions of clinical notes, claims, and payer-policy documents every month. Youll engineer autonomous pipelines that parse clinical records and translate into codes, provide documentation improvement parameters, and even solve for denials with autonomous calling if needed; Tackle the deep-domain challenges that make clinical and RCM AI one of the most rewarding problems in tech. Senior AI/ML Engineer- Job Overview We are hiring a Senior AI/ML Engineer to own the end-to-end applied LLM, retrieval, and evaluation layer of our healthcare AI platform. You will build production systems that automate mid- and end-revenue cycle workflows for US healthcare spanning coding, claim edits, denials triage, appeal generation, and payer-rule reasoning. This is a production engineering role (not research) focused on building scalable, auditable, and cost-efficient LLM systems in a regulated healthcare environment What Youll Own 1. Self-Hosted LLM Infrastructure Deploy, fine-tune, and operate open-source models (Llama, Qwen, MedGemma, and successors) as our primary inference stack Work with vLLM / SGLang / TensorRT-LLM for serving at scale, with disciplined attention to throughput, tail latency, batching, KV-cache, and GPU economics Own fine-tuning workflows end-to-end (SFT, LoRA, QLoRA, DPO) on clinical notes, claims, and payer-rule data Optimize GPU usage, latency, batching, and cost; make build-vs-buy and hosted-vs-self-hosted trade-offs explicit and measured 2. Knowledge Graphs & Embedding-Based Retrieval Design and maintain the knowledge graph encoding ICD-10-CM, CPT, HCPCS, modifiers, HCC, NCCI edits, LCD/NCD policies, and payer-specific rules and the relationships between them Build embedding-based retrieval over clinical notes, historical claims, denial reasons, and payer-policy corpora including chunking, embedding model selection, hybrid search, and reranking Combine graph traversal and dense retrieval so every coded line, scrubbed edit, and appeal response is grounded in auditable evidence Own ingestion, versioning, and quality of underlying knowledge sources (CMS, AHA, AMA, NCCI, payer bulletins) 3. Evaluation & Monitoring Build continuous evaluation pipelines that gate every model, prompt, retrieval, and graph change before production Run offline eval suites grounded in coder- and biller-validated labels; use LLM-as-judge where appropriate, calibrated against human ground truth Monitor drift, hallucinations, regressions, and output quality in production; operate shadow-mode rollouts and per-cohort accuracy tracking (specialty, payer, chart type) Track business metrics: chart-level and opportunity-level coding accuracy, denial rate impact, clean-claim rate, cost per chart, and end-to-end latency 4. LLM Systems & Prompt Engineering Design prompts and context pipelines for coding (CPT, ICD, HCC, E/M), claim edits, denial classification, and appeal drafting Implement structured outputs (JSON, function calling, constrained decoding) on top of the self-hosted stack Apply RAG over medical coding standards (CMS, ICD-10, AHA, NCCI) and payer policies, grounded in the knowledge graph and embedding stores Treat prompts as a thin, well-versioned, well-evaluated layer never the load-bearing piece 5. Agentic Workflows & Tooling MCP Build MCP servers for internal tools: code lookup, NCCI / rule checks, payer logic, eligibility, denial classification Design multi-step agent workflows with audit trails and human-in-the-loop checkpoints for coder, biller, and AR-analyst review Define deterministic vs. LLM-based tool boundaries for reliability reliability comes from knowing which is which What Were Looking For Must-Have 5+ years in ML/AI engineering, including 6+ months in production LLM systems Hands-on experience deploying and operating self-hosted LLMs (vLLM, SGLang, TensorRT-LLM, or equivalent) Strong experience designing embedding-based retrieval and/or knowledge graphs for grounded LLM applications Demonstrated ownership of evaluation infrastructure offline benchmarks, online monitoring, drift and regression detection Strong Python + PyTorch + Hugging Face experience Production experience with monitoring, incidents, and system ownership Strongly Preferred Fine-tuning experience (SFT, LoRA, QLoRA, DPO) on domain-specific corpora Experience .