Forward Deployed Software Engineer - US / NYC

The Senior Forward Deployed AI Engineer, Data Platform & Agents bridges the gap between complex enterprise data estates (Databricks, Lakehouses, vector stores) and frontier agentic AI execution (Anthropic Claude, multi-agent frameworks, custom MCP servers). Sitting embedded directly with client CDOs, VPs of Data, and engineering leadership, you will translate messy operational workflows and fragmented data sources into production-grade multi-agent platforms, real-time context retrieval systems, and LLM evaluation architectures.

This is a high-agency “builder-consultant” role for engineers who care as much about stakeholder adoption and runtime business impact as they do about model accuracy, protocol design, and low-latency data execution., * Agentic Tooling & MCP Architecture: Architect and deploy custom data interfaces and tool servers using the Model Context Protocol (MCP), function calling, and structured JSON schemas-enabling autonomous AI agents to query, synthesize, and execute actions across 10+ disparate enterprise data sources safely.

• Enterprise Lakehouse & Retrieval Engineering: Build high-performance data pipelines, vector search, and hybrid RAG layers over Lakehouse environments (Databricks, PySpark, Delta Lake, Unity Catalog)-slashing retrieval times and optimizing CPU/GPU query runtimes over massive structured and unstructured datasets.
• Embedded Client Execution & PoCs: Sit side-by-side with client technical teams to run daily requirements discovery, lead architecture reviews, and rapidly ship zero-to-one production Proofs of Concept (PoCs) in weeks rather than months.
• LLM Evals, Observability & Guardrails: Implement robust evaluation frameworks (LLM evals, hallucination tracking, retrieval precision/recall) and observability pipelines to monitor non-deterministic agent behavior and guarantee reliability in strictly regulated enterprise environments (Financial Services, Healthcare).
• Enterprise Security & Data Isolation: Collaborate with client CISO and security governance teams to enforce strict access-control lists (ACLs), multi-tenant isolation, row/column-level permissions, and zero-data-leakage constraints.

• Agentic AI & Orchestration: Proven track record building and deploying production multi-agent systems, RAG platforms, and MCP servers using frameworks like LangGraph, AutoGen, or CrewAI.
• Data Engineering & Lakehouse Depth: 5+ years of experience with Databricks, PySpark, Delta Lake, SQL, and Python-with specific expertise optimizing high-QPS analytical queries and processing unstructured enterprise documents (PDFs, contracts, logs).
• Statistical Rigor & Model Evals: Strong foundation in quantitative methods, statistics, or applied ML (M.S./Ph.D. or equivalent industry experience) to evaluate model uncertainty, ground truth, and non-deterministic systems systematically.
• Client-Facing / Consultative Muscle: Demonstrated ability to run technical discovery calls, manage non-technical client stakeholders, handle technical pushback, and lead co-engineering efforts on-site.

• High Autonomy Culture: Sharp, high-agency teams with no middle-management layers or corporate theater.
• Frontier-Level L&D: Generous learning budget, dedicated research time, and unencumbered access to state-of-the-art AI tools.
• Top-Tier Benefits: Competitive pay with performance bonuses, comprehensive health coverage, generous PTO, flexible holidays, parental leave, and a paid company shutdown the last week of December.

Indicium AI was Anthropic’s first European launch partner, a Preferred Anthropic Partner, and sits among a handful of organizations globally trusted to deploy Claude at an enterprise scale. Named the 2026 Databricks Consulting Partner of the Year and backed by Databricks Ventures, we build and deploy production AI systems for complex enterprises, making AI a genuine competitive advantage.