You will be a Principal Forward Deployed AI Engineer in the company's Transformation Command Center, working directly inside core business units on their hardest operational problems, where the goals are often unclear and the stakes are high.
You will own the technical side end to end, designing and shipping production agent systems, enterprise retrieval-augmented generation (RAG) engines, full-stack microservices and decision systems into live operations.
You will combine strong engineering with a product focus, turning messy operational problems into production architectures, and you will run workloads on Domino, Airflow and Kubernetes.
The work also includes packaging custom builds into reusable SDKs, shared libraries and architecture playbooks that other teams across the company can use.
Responsibilities
Define the full architecture for embedded AI initiatives, covering multi-agent orchestration, state machines, latency and cost budgets, security boundaries and back-end integrations.
Build stateful agent workflows that follow the company's agentic governance rules, using LangGraph, Temporal or event-driven state machines, with strict controls for tool calls, execution sandboxing and deterministic fallbacks.
Design production RAG systems with hybrid search, re-ranking, metadata filtering and semantic caching, shaping and scaling the company's context architecture.
Deploy agent services and scheduled workflows to production runtimes such as Domino Data Lab, Apache Airflow, Kubernetes microservices and serverless compute on AWS, GCP or Azure.
Design offline and online evaluation frameworks, including LLM-as-a-judge, synthetic test generation, human review queues and live A/B rollouts, to track model drift, hallucination rates and task completion.
Turn custom deployments into reusable internal SDKs, API building blocks and UI components, mentor engineers across the organization and write technical documentation for patent disclosures.
Work alongside business unit leaders to map complex processes, cut through noise and identify where applied AI will deliver measurable gains.
Act as the technical lead in meetings with business unit Vice Presidents and Managing Directors, explaining how algorithms behave, where they fail and what performance trade-offs apply.
Requirements
You'll need to have:
A bachelor's degree in computer science, software engineering or a related technical field, or 4+ years of work experience.
6+ years of relevant experience, gained through work, military service, specialized training or a mix of these.
Extensive hands-on work with frontier LLMs such as Anthropic Claude, OpenAI and Google Gemini, and with agentic developer tools such as Claude Code, Cursor and Copilot, to deliver scalable solutions and business results.
6+ years of shipping production software, or proven experience deploying LLM applications, agentic workflows or autonomous tool-calling pipelines to production.
Strong skills in Python and modern TypeScript/Node.js for full-stack API integration and back-end orchestration.
Experience with Docker containers, API design and cloud platforms (AWS, GCP or Azure).
A track record of building offline evaluation benchmarks and online evaluation and observability pipelines that track latency, token costs, hallucinations and eval-driven development.
Proven experience deploying containerized services and data workflows on enterprise platforms such as Domino Data Lab, Apache Airflow, Kubernetes (EKS/GKE) and hyperscaler clouds.
Advanced full-stack experience in Python, SQL and modern JavaScript/TypeScript.
A track record of designing offline and online eval harnesses, guardrail frameworks and observability systems covering latency, token optimization and tracing.
Ability to scope ambiguous technical work on your own and present architecture designs to senior executives.
Ability to work directly with enterprise stakeholders, turn unclear business needs into technical architectures and lead delivery from start to finish.
Even better if you have one or more of the following:
A master's degree in computer science, machine learning, robotics or a related technical field.
Experience building and scaling central context platforms or enterprise search engines.
Prior forward-deployed engineering, solutions architecture or technical consulting experience at leading technology firms or hyperscalers.
A record of turning one-off operational builds into company-wide shared libraries, with technical publications or filed patents.
Domain knowledge in telecommunications, network capacity optimization, capital planning or large-scale customer care.
Deep expertise in retrieval and context architectures: production RAG pipelines, vector databases such as pgvector, Milvus or Pinecone, hybrid dense and sparse search, and dynamic context management.
You are encouraged to apply even if you do not meet every 'even better' item.
Conditions
This is a hybrid role with a defined work location that includes working from home and a minimum of 3 days per week in the office, set by your manager. You are responsible for following the hybrid work policy.
Scheduled weekly hours are 40.
The base salary range for this location on a full-time schedule is $120,500.00 - $231,000.00. Actual pay depends on location and confirmed job-related skills and experience.
This is an incentive-based position with the potential to earn more. For part-time roles, pay is adjusted to match the hours worked.
Benefits
Health and wellness coverage, including medical, dental, vision, short and long term disability, basic and supplemental life insurance, AD&D, identity theft protection, pet insurance and group home and auto insurance.
A matched 401(k) savings plan, up to 8 company paid holidays and up to 6 personal days each year, paid parental leave, adoption assistance and tuition assistance, plus other incentives.
New hires receive up to 15 days of vacation per year, which increases with service.
Depending on the role, premium pay such as overtime, shift differential, holiday pay and allowances may be available.
Part-time staff may be eligible for some of these benefits, depending on individual circumstances.