Senior AI Engineer - GenAI Solutions

Responsibilities:

• Design and build production GenAI applications, LLM copilots, RAG pipelines, agentic workflows, and AI-native UX on our 2026 stack.
• Engineer retrieval end-to-end: chunking, embeddings, vector stores, retrieval, re-ranking, and hybrid search, and debug retrieval quality when it matters.
• Build agents and multi-step workflows with LangGraph and LangChain; integrate MCP servers and agent SDKs for tool access.
• Augment enterprise platforms Clarity PPM, Medallia, ServiceNow, Salesforce with predictive insight, intelligent automation, and conversational UX.
• Deploy to AWS Bedrock, Azure AI Foundry, or GCP Vertex AI with solid CI/CD, observability, and guardrails.
• Build evaluation harnesses, wire up LLM observability, and bake in responsible-AI basics: PII handling, guardrails, and audit trails.

Consult and Deliver:

• Run client discovery workshops and translate business problems into scoped, ROI-backed AI use cases.
• Demo working software to technical and business stakeholders, and defend your design decisions and tradeoffs.
• Tie every engagement to a measurable KPI: revenue lift, cost-to-serve, cycle time, adoption, or churn.

Requirements:

• 4-7 years software engineering: You've shipped and maintained real production systems, not just prototypes.
• Python mastery: Primary language for our AI services. Clean, testable, production Python (FastAPI or similar).
• Production GenAI experience: You've shipped at least one real LLM-powered feature to users: prompt design, structured outputs, tool/function calling, streaming.
• RAG engineering: Chunking, embeddings, vector DBs (Pinecone, Weaviate, pgvector, or Qdrant), retrieval, re-ranking, hybrid search. You can diagnose bad retrieval.
• Agentic systems: Built agents or workflows with LangGraph, LangChain, or LlamaIndex. You understand state, tools, and routing and what MCP and agent SDKs are for.
• Full-stack capability: Comfortable building AI-native UX with TypeScript / React / Next.js enough to ship a working product, not just an API.
• Cloud AI deployment Hands-on with at least one of AWS Bedrock, Azure AI Foundry, or GCP Vertex AI. Docker, CI/CD, and APIs shipped to production.
• Evaluation and trust: You build eval harnesses, use LLM observability (LangSmith, Langfuse, or Arize), and handle PII and guardrails responsibly.
• Client-facing communication- You can run a workshop, scope a use case, demo to stakeholders, and articulate ROI in plain business language. Fluent written and spoken English.