Architect and build end-to-end multi-agent AI systems with LangGraph: router / specialist / supervisor patterns, typed state, parallel retrieval, conditional routing and bounded retries
Design evidence-first solutions in which deterministic Python and SQL compute the facts and LLMs reason only over verified evidence, so outputs are grounded, cited and auditable
Build ingestion pipelines that turn unstructured sources (PowerPoint decks, PDFs, pipeline logs, Confluence pages, Jira tickets) into governed Bronze / Silver / Gold knowledge bases on Databricks and Unity Catalog
Implement advanced RAG: chunking and embedding pipelines, Databricks Vector Search, BM25 and hybrid retrieval, reciprocal rank fusion, cross-encoder reranking and metadata filtering
Build verification and guardrail layers (claim-level citation checks, numeric reconciliation, clarify / abstain behaviour) and LLM-as-a-judge evaluation for hallucination, completeness and actionability
Productionize AI workflows with Airflow and Databricks Jobs across dev, stage and prod, and ship AI applications as MCP servers, FastAPI services and Streamlit / React front ends on Databricks Apps or Kubernetes
Integrate Databricks Genie spaces, Agent APIs and Genie MCP endpoints into conversational analytics products, and set up governance and monitoring for AI agents
Enforce enterprise security and responsible AI: SSO, group-based access control that fails closed, service principals, secrets management and PII masking before data reaches a model
Lead a team of AI and data engineers: define the technical roadmap, break down work, review code and architecture, and mentor engineers on GenAI and agentic design patterns
Own delivery end to end: plan sprints, manage scope, risks and dependencies, and ensure releases meet quality, latency and cost targets
Act as the primary technical point of contact for client stakeholders: run discovery and solution workshops, present architectures and demos to business and technology leaders, and manage expectations through beta and production rollout
Identify new AI use cases with the client, shape proposals and case studies, and build reusable accelerators that can be replicated across clients
Architect and build end-to-end multi-agent AI systems with LangGraph: router / specialist / supervisor patterns, typed state, parallel retrieval, conditional routing and bounded retries
Design evidence-first solutions in which deterministic Python and SQL compute the facts and LLMs reason only over verified evidence, so outputs are grounded, cited and auditable
Build ingestion pipelines that turn unstructured sources (PowerPoint decks, PDFs, pipeline logs, Confluence pages, Jira tickets) into governed Bronze / Silver / Gold knowledge bases on Databricks and Unity Catalog
Implement advanced RAG: chunking and embedding pipelines, Databricks Vector Search, BM25 and hybrid retrieval, reciprocal rank fusion, cross-encoder reranking and metadata filtering
Build verification and guardrail layers (claim-level citation checks, numeric reconciliation, clarify / abstain behaviour) and LLM-as-a-judge evaluation for hallucination, completeness and actionability
Productionize AI workflows with Airflow and Databricks Jobs across dev, stage and prod, and ship AI applications as MCP servers, FastAPI services and Streamlit / React front ends on Databricks Apps or Kubernetes
Integrate Databricks Genie spaces, Agent APIs and Genie MCP endpoints into conversational analytics products, and set up governance and monitoring for AI agents
Enforce enterprise security and responsible AI: SSO, group-based access control that fails closed, service principals, secrets management and PII masking before data reaches a model
Lead a team of AI and data engineers: define the technical roadmap, break down work, review code and architecture, and mentor engineers on GenAI and agentic design patterns
Own delivery end to end: plan sprints, manage scope, risks and dependencies, and ensure releases meet quality, latency and cost targets
Act as the primary technical point of contact for client stakeholders: run discovery and solution workshops, present architectures and demos to business and technology leaders, and manage expectations through beta and production rollout
Identify new AI use cases with the client, shape proposals and case studies, and build reusable accelerators that can be replicated across clients
6+ years of experience in data engineering, AI/ML engineering or data platform development, with at least 2–3 years building LLM / Generative AI solutions that run in production
Proven, hands-on experience building agentic and multi-agent systems with LangGraph, LangChain or similar frameworks
Strong experience with RAG architectures: chunking strategies, embeddings, vector databases, hybrid search and reranking
Strong programming skills in Python (modular, typed, tested) and advanced SQL
Deep hands-on experience with Databricks: Apache Spark, Delta Lake, Unity Catalog, Model Serving / AI Gateway, Vector Search, Databricks SQL and Genie
Experience integrating LLMs through APIs (Anthropic Claude, OpenAI, Azure OpenAI, etc.), including structured outputs, tool / function calling and schema validation
Experience with LLM evaluation, guardrails and grounding techniques (LLM-as-a-judge, human-in-the-loop review, citation and numeric verification)
Experience with workflow orchestration (Airflow) and with building and deploying APIs and apps (FastAPI, Streamlit / React, Docker, Kubernetes)
Track record of leading and mentoring engineering teams of 3–8 people and owning delivery of complex AI projects
Strong client-facing skills: comfortable leading workshops, presenting to senior stakeholders, managing expectations and turning ambiguous business problems into clear solution roadmaps
Strong understanding of responsible AI: explainability, data privacy, access control and security in enterprise AI systems
Experience working on cloud-based AI solutions in AWS, Azure or GCP environments