AI Model Engineer

Signature Performance, Inc.

23h

About The Position

You are a person who is passionate about serving as a core technical leader within the Innovation, Research & Development (IR&D) department. We need someone who is responsible for architecting, optimizing, and operationalizing advanced AI systems that power Signature Performance's WorkflowXM platform, agentic workflow automation tools, and AI-enabled healthcare administrative solutions. In the role of AI Model Engineer, you will design, fine-tune, evaluate, and deploy large language models (LLMs), multimodal models, retrieval systems, and agentic frameworks that support high-volume medical coding, billing, claims processing, health insurance discovery, eligibility verification, customer service support, EDI ingestion, and enterprise automation.

Requirements

Bachelor's or Master's degree in Computer Science, AI/ML, Engineering, Applied Mathematics, or related field (or equivalent experience).
3--7+ years of hands-on experience designing, training, or deploying machine learning or deep learning models.
2+ years of direct experience with large language models, transformer architectures, or agentic AI systems.
Strong proficiency with Python, PyTorch, HuggingFace Transformers, LangChain/LlamaIndex, and vLLM or similar inference engines.
Expertise in fine-tuning, PEFT/LoRA tuning, prompt engineering, and RAG architectures.
Experience deploying models using Docker, Kubernetes, GPU/TPU compute environments, and model-serving frameworks.
Hands-on experience with vector databases (e.g., Qdrant), embeddings, and semantic search.
Understanding of NER, NLP, document processing, and multimodal ingestion (audio, text, PDF, EDI).
Ability to build agent-based systems with tool-calling, planning, verification loops, and chain-of-thought scaffolding.
U.S. Citizenship or naturalized citizenship is required for this position.
All work on all positions at Signature Performance must be completed in the continental United States, Alaska, or Hawaii.

Nice To Haves

Experience with health insurance, payer portals, COB/OHI logic, medical billing, claims processing, medical coding, or EDI/HL7/FHIR standards.
Understanding of HIPAA, PHI/PII handling, and secure AI practices.
Strong problem-solving and analytical mindset.
Ability to work in fast-paced environments with ambiguous requirements.
Clear communication skills for collaborating with technical and non-technical stakeholders.
Passion for innovation and transforming administrative healthcare through AI.

Responsibilities

Design, fine-tune, distill, and optimize LLMs and multimodal models to support WorkflowXM and other enterprise AI initiatives.
Implement PEFT/LoRA/Q-LoRA fine-tuning strategies for domain-specific healthcare tasks.
Develop retrieval-augmented generation (RAG) pipelines using vector databases (e.g., Qdrant).
Architect and evaluate model prompting frameworks, system prompts, and agentic prompt templates.
Optimize inference speed, accuracy, grounding, and reliability across on-prem and cloud environments.
Build and maintain tool-using AI agents, orchestrators, governance watchdog agents, and multi-agent workflows.
Implement memory systems, shared context, scratchpads, tool-calling logic, and verification loops.
Integrate AI agents with internal systems, including MCP servers, WorkflowXM modules, backend microservices, and data ingestion pipelines.
Create reusable "AI capability modules" for insurance discovery, claims QA, coding assistance, EDI parsing, and document understanding.
Develop evaluation datasets, synthetic tests, and red-teaming harnesses for model quality assurance.
Implement guardrails for PHI/PII protection, hallucination minimization, and HIPAA-compliant behavior.
Monitor model drift, regressions, latency metrics, and accuracy degradation.
Work closely with IR&D governance and security partners to align models with Data Ethics & AI Governance Policy (CCP-046).
Collaborate with AI Systems Engineers and DevOps to deploy, containerize, and scale models in production.
Support GPU/accelerator utilization strategy and model serving frameworks (vLLM, Ollama, HuggingFace, TensorRT-LLM).
Assist in designing and maintaining model registries and versioned deployment systems.
Ensure high-throughput, high-availability inference services for WorkflowXM and enterprise applications.
Partner with analysts, developers, QA, clinical SMEs, and business stakeholders to translate operational needs into AI capabilities.
Contribute to AI-driven modernization of claims processing, revenue cycle workflows, EHR optimization, and payer service operations.
Support rapid prototyping, experimentation, and innovation aligned with IR&D's strategic priorities.
Document architectural approaches, model behaviors, and best practices for internal knowledge growth