AI Architect / Forward Deployed Engineer

Bachelor’s degree in computer science, Engineering, Data Science, or a related technical field. Master's degree (M.Tech / MS / MBA) preferred.

Experience

• 10+ years of progressive experience in software/AI engineering, cloud architecture, or forward deployed/solution engineering roles.

• Deep cloud experience is a must — proven hands-on architecture and deployment across at least one major hyperscaler (Azure / AWS / GCP), including production-grade, secure, scalable systems in client environments.

• Deep AI Agents experience is a must — demonstrated design and delivery of Agentic AI / multi-agent and Generative AI systems (LLMs, RAG, orchestration frameworks) in production.

• Deep healthcare experience is a must — strong understanding of U.S. healthcare payer operations (Prior Auth, UM, Care Management, or Claims).

• Proven track record of personally taking AI solutions from POC to live production—hands-on, not purely advisory.

Technical Skills

• Hands-on and architectural expertise in LLMs, embeddings, vector search, prompt engineering, and RAG pipelines.

• Strong coding ability (Python and related AI/ML stacks) with the ability to ship production-quality solutions directly.

• Proficiency with cloud AI platforms: Azure OpenAI, AWS Bedrock (Claude/Sonnet), GCP Vertex AI.

• Expertise with agent orchestration frameworks: LangChain, LangGraph, CrewAI, AutoGen, or equivalent.

• Strong understanding of MCP (Model Context Protocol), A2A protocols, and multi-agent system design.

• Infrastructure-as-code, Docker/Kubernetes, CI/CD, MLOps/LLMOps, and GPU capacity planning.

• Secure API design, OAuth2/JWT, enterprise integration patterns, and healthcare data standards (HL7, FHIR, X12 EDI 837/835/270/271/276/277).

Domain Skills

• Deep understanding of payer clinical and claims workflows and integration with platforms such as QNXT, Facets, CareRadius.

• Familiarity with clinical criteria sources (MCG, InterQual) and CMS regulatory frameworks (including CMS-0057-F).

Working Style

• Comfortable being embedded on-site/in-client-environments with significant travel and direct, day-to-day collaboration with client technical teams.

• Bias for action, strong ownership, and ability to operate autonomously in ambiguous, fast-changing conditions.

• Ability to lead distributed US + offshore engineering support while remaining hands-on.

────────────────────────────────────────────

Preferred Qualifications

• Experience building AI Centers of Excellence or reusable agent frameworks/accelerators for healthcare.

• Patent holder, published researcher, or recognized contributor in AI/healthcare innovation.

• Experience with Databricks, Snowflake, or similar data platforms in a healthcare context.

• Cloud/AI certifications (Azure Solutions Architect, AWS Solutions Architect Professional, GCP Professional Architect, or equivalent).

1. AI Solution & Reference Architecture Ownership

• Own the end-to-end technical architecture for AI-powered healthcare solutions—covering agent orchestration, LLM serving, retrieval pipelines, data ingestion, and integration with legacy payer systems.

• Define and maintain reusable reference architectures, design patterns, and architectural guardrails for Agentic AI solutions that can be rapidly adapted across client deployments.

• Establish and govern AI Dev/QA/Prod environments, CI/CD pipelines, and MLOps/LLMOps practices, doubling as the AIOps architect for the team.

2. Forward Deployment & Hands-On Delivery

• Embed within client environments to understand real workflows, data, constraints, and edge cases firsthand translating them directly into working AI solutions.

• Personally build, configure, and deploy solutions inside client infrastructure ("bring your own subscription" / client-hosted), not just design them.

• Operate in fast, iterative feedback loops—prototype, demo, gather input, and harden to production at speed.

• Troubleshoot live deployments, resolve integration and performance issues, and ensure solutions run reliably in the client's operational context.

• Capture reusable learnings, accelerators, and IP from each deployment to shorten time-to-value on the next.

3. Cloud Architecture & Engineering

• Architect and implement secure, scalable, multi-cloud and hybrid deployments across Azure, AWS, and GCP—including within client-controlled subscriptions and networks.

• Build infrastructure-as-code, containerized (Docker/Kubernetes) deployments, GPU capacity planning, and cost/latency optimization for LLM workloads.

• Implement AI gateway design, PHI de-identification layers, and enterprise integration patterns with payer platforms and data stores.

• Ensure deployments meet stress/throughput targets (e.g., high-volume FHIR record processing) with resilient, observable systems.

4. AI Agents & Generative AI Design

• Architect and build multi-agent systems—policy decision-tree agents, clinical review agents, data-gathering agents, and IDP agents—that replicate expert human judgment with human-in-the-loop governance.

• Set standards for prompt engineering, RAG pipelines, fine-tuning/domain adaptation, embeddings, and vector search.

• Build agent execution layers for legacy system navigation (mainframe, Onbase/ECM, CareRadius, Salesforce, Epic) using MCP servers and browser/desktop automation.

• Champion explainability-first, audit-ready AI design suitable for CMS audits and payer compliance reviews.

5. Healthcare Domain Integration

• Apply deep knowledge of payer workflows—Prior Authorization, Utilization Management, Care Management, Claims, and Provider operations—to translate operational challenges into deployed AI solutions.

• Integrate AI solutions with clinical criteria sources (MCG, InterQual, CMS medical policies) and payer platforms (QNXT, Facets, CareRadius) via API-first, standards-based design (HL7, FHIR, X12 EDI).

• Support alignment with regulatory drivers such as CMS-0057-F interoperability/prior-auth mandates.

6. Governance, Security & Compliance

• Champion privacy-first design, data anonymization, and compliance with HIPAA and PHI/PII handling requirements.

• Establish governance-grade AI controls—model monitoring, bias detection, drift management, and human-in-the-loop overrides—within each deployed environment.