- Company Name
- Tror - AI for everyone
- Job Title
- AI/ML Architect
- Job Description
-
Job Title: AI/ML Architect
Role Summary
Lead the design and deployment of scalable, observable, and compliant AI/ML solutions that integrate seamlessly with clinical workflows and hospital data platforms. Own the full model lifecycle from development to drift monitoring, with a focus on HIPAA‑regulated environments and real‑time inference.
Expectations
- 8+ years of experience in AI/ML or data architecture roles.
- Proven track record of designing end‑to‑end ML pipelines and MLOps operations in cloud environments.
- Strong understanding of clinical data standards (FHIR, HL7, SMART‑on‑FHIR) and experience with enterprise EHR platforms (e.g., Epic).
- Ability to architect and secure GenAI workloads, including LLMs, embeddings, and agentic systems.
Key Responsibilities
- Design and implement end‑to‑end ML pipelines using Airflow, MLflow, Vertex AI, or Azure ML.
- Build and deploy containerized inference services on Kubernetes (K8s) with Docker, ensuring observability through AppDynamics or similar.
- Own the model lifecycle: versioning, rollback, shadow testing, and drift monitoring.
- Architect GenAI pipelines (LLM inference, embeddings, RAG) with tools such as Vertex AI, Azure OpenAI, LangChain, and vector databases (FAISS, Pinecone, Weaviate, ChromaDB).
- Implement governance guardrails for agentic systems, including prompt injection protection, moderation, and output fallback.
- Integrate real‑time inference services with Epic via FHIR APIs, ensuring end‑to‑end audit trails and access controls.
- Enforce HIPAA‑compliant encryption, access controls, and audit logging across all services.
- Collaborate with cross‑functional teams to define and enforce MLOps best practices and compliance standards.
Required Skills
- Kubernetes, Docker, and container orchestration expertise.
- Deep familiarity with GCP Vertex AI, Azure ML, and Snowflake.
- Proficiency in building and monitoring ML pipelines (Airflow, MLflow).
- Experience with HIPAA‑regulated, real‑time model deployment and security controls.
- Knowledge of clinical data standards (FHIR, HL7, SMART‑on‑FHIR).
- Strong scripting/automation skills (Python, Bash, Terraform).
- Familiarity with GenAI tools (LLMs, embeddings), LangChain, and vector databases.
Required Education & Certifications
- Bachelor’s or Master’s degree in Computer Science, Data Science, or related field.
- Relevant certifications, such as Certified Kubernetes Administrator (CKA), GCP Professional AI Engineer, or Azure AI Engineer Associate, are preferred.