Define guardrails for agent behavior, including least-privilege access, human-in-the-loop checkpoints, action logging, and escalation paths.
Establish evaluation and observability standards for agents: task success, safety, reliability, cost per task, and regression testing before release.
Drive interoperability standards such as MCP and open agent-to-agent protocols, and own the agent operating model: ownership, lifecycle, and decommissioning.
Own the model strategy across frontier models such as Anthropic’s Claude, OpenAI and Google Gemini, and open-weight models such as Llama, Mistral and Qwen, matching each workload to the right model on quality, cost, latency and data-residency needs.
Stand up and govern an enterprise AI gateway that centralizes model access, routing, fallback, rate limiting, cost tracking and policy enforcement across every team and agent.
Decide when to self-host open-weight models (via vLLM, Ollama or managed endpoints) and when to use hosted APIs, and run the evaluation that backs each choice.
AI data foundation and trust
Define and evolve the reference architecture across the knowledge graph, semantic layer, vector retrieval, and hosted trust layer that grounds every AI agent.
Own the trust layer charter: classification, masking, lineage, and policy enforcement for data that lacks native role-based access control.
Lead batch and real-time pipelines that move source data into AI-ready form for models and agents.
Set data modeling standards, canonical entities, and shared business definitions, and build versioned feature and context stores with freshness, latency, and access SLAs.
Establish SLOs for retrieval quality, data freshness, and trust-layer policy compliance, backed by data quality checks, drift monitoring, and lineage so every agent action traces to its data.
Balance performance and cost across storage, query, and compute for training, retrieval, and inference.
Team and thought leadership
Mentor, and grow a team of AI and data engineering professionals as a player-coach, with a culture of exploration and excellence.
Lead async, multi-geo collaboration norms across the US and India, and defend team scope against feature-factory pressure.
Serve as "Customer Zero" for Boomi's low-code generative AI orchestration platform, and share practical lessons with customers and partners.
Stay ahead of AI trends, and work with Boomi Product & Engineering and external communities to test ideas and sharpen the AI strategy.
The Experience You Bring
10+ years across data engineering, analytics, and AI systems, including 3+ years managing senior IC engineers.
A track record leading enterprise-scale AI initiatives that turned large business transformations into measurable outcomes.
Hands-on experience building production solutions on Claude, including the Claude API, Claude Agent SDK, tool use and prompt caching, plus at least one other frontier model family.
Experience deploying, fine-tuning or serving open-weight models in production, and comparing them against hosted models on quality and total cost.
Experience designing or operating an AI gateway or LLM proxy layer (such as LiteLLM, Portkey, Kong AI Gateway, Cloudflare AI Gateway or AWS Bedrock) for multi-model routing, observability and governance.
Has personally architected at least one of: a production knowledge graph, a semantic layer, or a RAG retrieval system at enterprise scale, and stays close to the code as a player-coach.
Experience setting architecture for agentic or multi-agent AI systems, including guardrails, evaluation, and observability.
Sponsorship data shows employers' past filings in public US government records (DOL, USCIS, E-Verify) and is not a promise that any job will be sponsored. Immigration rules change. Verify your situation with your DSO or an immigration attorney.