VP – Distinguished Engineer of Generative AI Engineering

• Design and own the end-to-end GenAI platform: context layer, data layer, model serving, agent frameworks, and evaluation pipelines. Architecture decisions here must hold at enterprise scale, across multiple product lines, manufacturing sites, and a growing engineering organization. This means designing for reliability, cost efficiency, and extensibility from day one, not retrofitting those properties after the fact.

•

• 15+ years of engineering experience, including 5+ years shipping production GenAI systems and 7+ years in a senior technical leadership role. Candidates are expected to have work that is recognized outside their own organization.

• BS required. MS or PhD in Computer Science, Machine Learning, Robotics, Electrical Engineering, Mechanical Engineering or a related field preferred. Exceptional track record beats pedigree. Candidates without advanced degrees who have published, built widely-adopted systems, or otherwise demonstrated research-level thinking through their work will be evaluated accordingly.

• Deep hands-on experience with LLM APIs (OpenAI, Anthropic, Gemini) and open-source model deployment at production scale.Vector databases, agent orchestration frameworks (LangChain, LlamaIndex, or equivalent), and prompt engineering for reliability, not just capability.

• Fine-tuning and training purpose-built models for domain-specific tasks, including the tradeoff analysis between hosted APIs, open-source models, and custom training across cost, latency, and IP dimensions.

• Evaluation frameworks, guardrails, and observability pipelines for GenAI in production: the ability to know whether a system is working, not just whether it is running.

• Experience designing and deploying multi-agent systems in production, including coordination protocols, failure isolation, and human-in-the-loop escalation paths.

• Candidates with this background will move faster in the role; those without it should expect to develop it on the job.Deploying AI systems that operate on physical hardware: robots, CNC machines, quality inspection systems, assembly line sensors, or equivalent industrial environments.

• Computer vision systems for manufacturing quality assurance: defect detection, dimensional inspection, or process monitoring at production-line throughput.

• Closed-loop learning systems where model behavior is updated based on physical world outcomes, including approaches to safe online learning in environments where errors have physical consequences.

• Robot-to-model feedback architectures: designing systems where robotic process data improves model performance and model outputs improve robotic process control.

• Familiarity with physical simulation environments (Isaac Sim, MuJoCo, or similar) for model development and validation before factory deployment.

• Sensor fusion and multimodal data handling for environments with structured sensor data, unstructured imaging, and natural language interfaces operating simultaneously.

• Designing real-time inference pipelines with hard latency constraints and failure consequences, including edge deployment architectures that operate under intermittent connectivity.

• Data architecture for AI at scale: modern data stacks (Snowflake, Databricks, dbt), vector stores, and streaming data pipelines that handle both batch and real-time model consumption.Cloud infrastructure across AWS, GCP, or Azure at the architecture level, including cost modeling, capacity planning, and the tradeoffs between managed AI services and self-hosted infrastructure.

•

Designing and owning AI platforms that serve enterprise-scale internal usage: thousands of concurrent users, dozens of integrated systems, and SLA requirements that real operations depend on.

Multi-tenant model serving architectures with cost isolation, quota management, and workload prioritization across competing internal consumers.Cloud infrastructure across AWS, GCP, or Azure at the architecture level, including cost modeling, capacity planning, and the tradeoffs between managed AI services and self-hosted infrastructure.Security and compliance architecture for AI systems handling proprietary manufacturing IP, including data residency, access controls, and audit trails.

Platform engineering practices that make GenAI accessible and reliable for internal teams who are not AI practitioners: SDKs, abstraction layers, and self-service tooling.