Founding VP of Engineering (ADK)

Pelico•Boston, MA

1d•Remote

About The Position

Pelico is seeking a Founding VP of Engineering to lead the development of a new agentic orchestration platform. This is a hands-on leadership role with significant ownership, responsible for setting technical direction, building a team, and delivering a platform that allows external companies to autonomously build and deploy their own agents. The role benefits from existing company assets such as data, event dependency graphs, a simulation engine, and next-best-action algorithms, and works with flagship manufacturers as design partners. The goal is to create a platform for manufacturers and partners to autonomously build and deploy their own agents, addressing the critical need for timely responses in supply chain disruptions.

Requirements

Experience building and operating production multi-agent systems, including orchestration runtimes, tool and MCP surfaces, agent harnesses, guardrails, and model selection.
Experience taking an early product from its first design-partner customer to revenue in a demanding or regulated industry (aerospace, defense, finance, healthcare), managing trust and the sales cycle from demo to deal.
Demonstrable AI fluency, including shipping AI-agent products with robust evaluation and observability (regression and drift detection, cost and latency budgets).
Daily experience building with fleets of AI coding agents and understanding of automation versus human intervention.
Ability to write and review production code daily and lead senior engineers.
A platform builder's instinct for turning internal engines into clean surfaces for other teams.

Nice To Haves

Experience in supply chain, discrete manufacturing, or operations (planning, parts shortages, inventory).
Sufficient optimization or operations research literacy to collaborate with an OR team.
Experience in aerospace, defense, or other regulated industries (e.g., ITAR).

Responsibilities

Build the platform: Design and ship the system manufacturers use to author, run, observe, and govern their agents, including the canvas, tool catalogue, model selection, and the full loop.
Turn the twin into an environment agents can operate in: Open up the deterministic engine so an agent can read the plan, take an action, watch the consequences propagate, and be scored on the impact to the whole factory.
Solve the multi-agent problem: Measure each agent's true contribution and prevent agents from interfering with each other's work.
Earn the enterprise's trust: Ensure every agent action is reproducible, auditable, and explainable to the standards required by regulated aerospace and defense IT departments, with appropriate human oversight.
Find product-market fit: Differentiate between current capabilities and future vision, decide platform scope versus individual product scope, and develop a compelling demo and commercial story for enterprise deals.
Set the bar and build the team: Write specifications and production code, leverage AI coding agents in the development process, and hire and grow a team of engineers.