Senior AI Engineer, Forward Deployed Engineering

NVIDIA•Santa Clara, CA

About The Position

Join NVIDIA Operations’ AI & Automation Supply Chain Business Transformations team to build decision-intelligence applications for the supply chain using Palantir Foundry and AIP. This Senior AI Engineer and Forward Deployed Engineer works directly with planners and supply-chain experts to turn decision logic into production applications, including governed data pipelines, Ontology models, analytical functions, operational interfaces, and AI agents. The applications should be trusted, tested, and used in daily operations. The role combines Palantir engineering, applied AI, and supply-chain transformation. It involves delivering and owning production applications after go-live, translating business rules and operational requirements into maintainable software, testing and evaluating AI-enabled workflows, establishing appropriate human oversight, and measuring operational outcomes. The position also leads design and integration work and supports user acceptance testing and hypercare.

Requirements

Demonstrated hands-on production delivery on Palantir Foundry and AIP, including Python or Pipeline Builder transforms, Ontology design (object, link, and action types), Workshop or OSDK applications, AIP Logic and Functions, and Data Connection sources, with ownership after go-live.
Working knowledge of Foundry governance and release management, including projects and roles, markings, branching, and promotion across development, UAT, and production environments.
Bachelor’s degree in Computer Science or a related field, or equivalent experience.
12+ years of demonstrable experience in software architecture, systems design, or enterprise platform engineering.
Strong Python and SQL development skills.
Experience integrating enterprise systems such as ERP, planning, or ITSM through APIs, data connectors, or custom services.
Ability to implement complex business rules and explain how their correctness was established.
Strong automated testing, debugging, and release practices, including validation of AI-assisted development and LLM-based workflows.
Practical experience integrating and evaluating language models, including open models such as NVIDIA Nemotron.
Ability to work directly with operational users, resolve ambiguous requirements, and deliver maintainable software.
Excellent collaboration and communication skills, including the ability to explain complex designs clearly and build trusted partnerships across technical and business audiences.

Nice To Haves

Supply-chain, manufacturing, planning, or operations research experience.
Experience with constraint optimization, simulation, and scenario analysis.
TypeScript development and advanced Palantir application development.
Experience with agentic frameworks and patterns such as RAG pipelines, multi-agent orchestration, and vector search, and with NVIDIA NeMo Guardrails or GPU-optimized inference stacks.
Hands-on experience with containerized deployments, Kubernetes, and hybrid cloud/on-prem environments.
Evidence of measurable improvements from systems used in daily operations.

Responsibilities

Build and own production supply-chain applications on Palantir Foundry and AIP, including data pipelines, Ontology object, link, and action types, analytical functions, Workshop and OSDK interfaces, and controlled actions.
Work with supply-chain planners and experts to turn business rules, constraints, and exception procedures into documented, versioned, and testable software.
Combine analytical methods, optimization, and language models to support operational decisions, choosing approaches that meet correctness, reliability, and performance requirements.
Own automated testing and validation across application code, data pipelines, business rules, and agent workflows.
Establish acceptance criteria with domain experts and enforce checks before release.
Review AI-generated code, validate assumptions, and verify behavior against independently defined expected results.
Build evaluation datasets from approved scenarios and historical decisions.
Test edge cases, repeatability, permissions, failure recovery, and changes to models, prompts, tools, and data.
Define agent autonomy boundaries, human-in-the-loop workflows, and escalation strategies for enterprise supply-chain use cases built on AIP.
Measure operational results such as decision cycle time, manual effort, exception resolution, and decision quality; capture user corrections and outcomes to guide improvements.
Capture design requirements, translate them into user stories, lead technical design reviews and documentation, drive end-to-end integration testing, and support UAT and hypercare.