Role Summary
This is a senior, hands-on AI architecture role focused on production-grade AI systems rather than proofs-of-concept. You will act as the technical voice for enterprise AI, guiding use case validation, solution architecture, stakeholder engagement, AI governance, and production deployments, especially in regulated environments like healthcare.
Key Responsibilities
1. Use Case Validation & Business Case Development
- Evaluate incoming AI requests for feasibility, ROI, data availability, and achievable accuracy.
- Make clear go / pause / abandon decisions based on technical/business analysis.
- Identify when simpler ML, rules, or off-the-shelf solutions outperform custom GenAI investment.
- Standardize AI request intake and scoping with product managers.
2. Solution Architecture & Technical Strategy
- Recommend appropriate AI approaches: RAG pipelines, agent orchestration, fine-tuning, prompt engineering, or hybrid solutions.
- Design PoCs with pre-defined success criteria.
- Maintain reference architectures for recurring AI patterns, microservices, and enterprise-grade workflows.
- Evaluate AI frameworks, models, and infrastructure under enterprise and healthcare constraints.
3. Business Stakeholder Partnership
- Work directly with operations, product, and clinical leaders throughout the AI lifecycle.
- Educate stakeholders on AI's probabilistic nature and establish realistic success metrics.
- Present architecture decisions and recommendations to non-technical audiences.
4. Enterprise AI Governance, Safety & Evaluation
- Enforce guardrails: PII redaction, hallucination mitigation, input/output filtering.
- Implement evaluation frameworks: automated regression testing, human-in-the-loop validation, ground-truth datasets.
- Ensure explainability, bias testing, audit logging, and compliance for healthcare AI systems.
- Enable distributed tracing for complex multi-agent workflows.
5. Technical Leadership & Deployment
- Lead architecture direction, code reviews, and DevOps guidance for GenAI and ML engineers.
- Guide backlog prioritization and technical trade-offs for product managers.
- Architect end-to-end Azure deployments: AKS, Azure OpenAI, Azure AI Search, KONG API Gateway.
- Establish secure APIs, secrets management, model versioning, and LLMOps practices (prompt versioning, A/B testing, drift detection, retraining triggers).
Required Qualifications
- 10+ years software engineering, 4+ years in AI/ML production systems.
- Hands-on experience building RAG pipelines, agentic AI, or LLM-based applications at scale.
- Proven ability to work directly with stakeholders and make go/no-go decisions.
- Deep expertise in enterprise AI governance and regulated environments.
- Strong Azure expertise: OpenAI Service, AI Search, AKS, ML, Key Vault.
- Production Kubernetes experience, deployment, autoscaling, observability.
- Secure API design: gateways, OAuth, secrets management.
Strongly Preferred
- LLM orchestration frameworks: LangChain, Semantic Kernel, CrewAI.
- RAG evaluation/observability tooling: RAGAS, DeepEval, LangSmith.
- ML background: embedding models, fine-tuning, retrieval optimization.
- Healthcare/insurance experience: HIPAA, PHI, data governance.
- Salesforce platform exposure: Service Cloud, Einstein AI, Agentforce, LWC development
Ideal Candidate Profile
- Has stopped projects that shouldn't proceed using data-driven judgment.
- Experienced in redirecting unrealistic expectations on AI accuracy with stakeholders.
- Personally designed healthcare AI guardrails and overseen production incidents.
- Operates with hands-on execution experience, not just advisory knowledge.