Senior AI/ML Engineer, Healthcare NLP & LLMs

Senior AI/ML Engineer

Artificial intelligence is central to Avail Health’s strategy for delivering high-quality, scalable care to Medicare-age populations. Our AI/ML team designs and deploys production models that drive clinical documentation intelligence, risk adjustment accuracy, predictive population health analytics, and intelligent care-gap identification. We operate at the intersection of applied machine learning, clinical NLP, and healthcare data standards—building systems that must be both technically rigorous and clinically trustworthy.

The Senior AI/ML Engineer is core member of Avail Health’s technology team, responsible for researching, designing, training, and deploying machine learning models and NLP systems that power our clinical operations, risk adjustment workflows, and care delivery platform. This role requires deep hands-on expertise in the full ML lifecycle—from data engineering and feature development through model training, evaluation, and production deployment—with a strong emphasis on healthcare NLP, large language model (LLM) fine-tuning, and responsible AI in regulated environments.

This position reports directly to the Chief Technology Officer (CTO) and works cross-functionally with Clinical Operations, Product, Analytics, Quality, and Revenue Cycle teams to translate clinical and business objectives into high-impact ML solutions.

This is a full-time, offshore [AF1] [JH2] position. The role is fully remote and requires a reliable high-speed internet connection and a dedicated workspace to support secure communications and handling of protected health information (PHI). Flexibility is required to accommodate regular collaboration with U.S.-based teams across multiple time zones, primarily Eastern Time.

This is a hands-on senior individual contributor role with deep ownership over Avail Health’s ML and NLP capabilities. The Senior AI/ML Engineer will lead model development end-to-end: from problem framing and dataset curation through architecture selection, training, evaluation, and production deployment.

The ideal candidate brings deep expertise in training and fine-tuning large language models and applying NLP to unstructured clinical text—such as encounter notes, care plans, and structured EHR data—to extract meaning, support coding accuracy, and surface actionable insights. Experience working with healthcare data standards (HIPAA, HL7/FHIR, HCC/RAF scoring, ICD-10) and building models under regulatory constraints is essential.

This individual will also contribute to ML infrastructure and MLOps practices, mentor peers on applied ML methodology, and help establish the standards and architecture that will define Avail Health’s AI capabilities as the organization scales.

Design, implement, and train ML and deep learning models across a range of supervised, semi-supervised, and self-supervised learning paradigms. Fine-tune large language models (e.g., GPT-class, BERT-family, open-weight instruction-tuned models) for healthcare-specific tasks including clinical NLP, document classification, information extraction, and summarization. Apply structured training methodologies including RLHF, PEFT (LoRA, QLoRA), and instruction tuning. Rigorously evaluate models using domain-appropriate benchmarks and clinical validation frameworks.

Build and maintain NLP pipelines for processing unstructured and semi-structured clinical text, including physician and nurse[AF3]  encounter notes, care assessments, prior authorization documents, and discharge summaries. Develop and deploy named entity recognition (NER) models for clinical entities (diagnoses, medications, procedures, symptoms), relation extraction pipelines, assertion classifiers, and negation/uncertainty detection. Align NLP outputs to clinical coding standards including ICD-10-CM, HCC categories, and SNOMED CT where applicable.

Develop predictive models targeting Medicare Advantage risk adjustment workflows, including HCC coding accuracy, prospective RAF scoring, care-gap identification, and HEDIS/STARS measure prediction. Integrate model outputs into clinical operations tooling to surface actionable insights for care teams and coders. Design feedback loops and active learning mechanisms to improve model performance continuously against real-world clinical outcomes.

Design and implement production AI systems that leverage LLMs as core reasoning components, including retrieval-augmented generation (RAG) pipelines, multi-step agentic workflows, and structured extraction tools operating over FHIR-native clinical data. Evaluate and select foundational models, embedding strategies, and vector store architectures appropriate to clinical use cases. Ensure LLM deployments meet requirements for explainability, auditability, and clinical safety in a regulated healthcare context.

Own data pipelines from clinical source systems (EHR, claims, FHIR APIs, ADT feeds) through feature engineering, dataset curation, and training/evaluation split construction. Build and maintain scalable data processing workflows using tools such as Spark, dbt, Airflow, or equivalent. Partner with data engineering to ensure ML feature stores are version-controlled, reproducible, and audit-ready.

Deploy and operate ML models in production using cloud-native MLOps patterns including containerized inference services, model registries, experiment tracking, drift monitoring, and automated retraining pipelines. Design for observability, latency SLAs, and graceful degradation. Maintain model documentation, versioning, and performance monitoring dashboards appropriate for a HIPAA-regulated environment.

Remain current with the applied ML and clinical NLP research landscape, evaluating and integrating relevant advances into production workflows. Contribute to and/or lead internal ML research initiatives. Mentor junior ML engineers and data scientists through technical reviews, pair work, and structured knowledge sharing. Communicate model behavior, limitations, and tradeoffs clearly to clinical, product, and executive stakeholders.

•      Bachelor’s degree in Computer Science, Machine Learning, Statistics, Computational Linguistics, or a related technical field; Master’s or Ph.D. preferred; or equivalent combination of education and professional experience

•      2+ years of professional experience in applied machine learning or AI engineering, with a demonstrated track record of deploying models to production

•      Deep hands-on experience training and fine-tuning large language models and transformer-based architectures (e.g., BERT, RoBERTa, T5, LLaMA, Mistral, GPT variants)

•      Strong NLP fundamentals: tokenization, embeddings, sequence labeling, text classification, relation extraction, coreference resolution, and generative modeling

•      Proficiency in Python and core ML/NLP libraries: PyTorch or TensorFlow/JAX, HuggingFace Transformers, spaCy or stanza, scikit-learn, NLTK

•      Experience with PEFT methods (LoRA, QLoRA, adapters), RLHF, and instruction tuning workflows

•      Hands-on experience building RAG pipelines, vector stores (e.g., Pinecone, Weaviate, pgvector), and embedding strategies for domain-specific retrieval

•      Solid grounding in ML evaluation methodology: train/dev/test splits, cross-validation, confidence calibration, bias/fairness auditing, and statistical significance testing

•      Experience with MLOps tooling: experiment tracking (MLflow, W&B), model registries, containerized inference (Docker, Kubernetes), and CI/CD pipelines for ML workflows

•      Working knowledge of HIPAA compliance requirements and data governance practices for handling PHI in ML systems

•      Excellent written and verbal communication skills in English, with demonstrated ability to explain model behavior and tradeoffs to non-technical stakeholders

•      Ability to maintain a minimum of 4 hours of daily overlap with U.S. Eastern Time business hours (8 a.m.–5 p.m. ET)

•      2+ years of experience applying NLP or ML to clinical text, EHR data, or healthcare claims data

•      Familiarity with healthcare NLP frameworks and resources

•      Experience with clinical coding standards: ICD-10-CM, CPT, HCC categories, SNOMED CT, RxNorm, or LOINC

•      Working knowledge of Medicare Advantage risk adjustment, HCC/RAF scoring mechanics, or HEDIS/STARS quality measures

•      Experience building or consuming FHIR R4 APIs and processing HL7 clinical data formats

•      Experience with clinical document understanding: section detection, note summarization, problem list generation, or CDI tooling

•      Familiarity with privacy-preserving ML techniques: federated learning, differential privacy, or de-identification pipelines for PHI

•      Published research, conference contributions, or open-source work in NLP, clinical AI, or applied ML

•      Experience in a startup or high-growth environment building ML systems and practices from the ground up

•      Prior experience working as part of a distributed, offshore, or globally remote engineering team