How Factory AI Does Forward Deployed Engineering

These are my notes from the Forward Deployed Engineering workshop at @aiDotEngineer on June 30, 2026.

Session: How Forward Deployed Engineering Is Done at @FactoryAI
Speaker: Eno Reyes
Company: Factory AI
X: @EnoReyes

What Factory AI Does

Factory provides building blocks for a "software factory": the implicit process every organization uses to build, evolve, and scale software.

A lot of Factory's work goes into understanding each customer's environment, workflows, and end-to-end story. The offering has also evolved away from pure human strategy toward more product-driven delivery.

Core AI Capability

The Factory framing was that deployed engineering means AI autonomously completing tasks without human intervention.

The key condition is verifiability. AI becomes much more competent when the "done" state is clear and externally checkable. Any problem should be framed as a set of verification systems.

The system works by pushing inference until task completion, given a clear goal and success criteria.

Problem Space and Use Cases

Factory targets complex, large-scale migrations.

The best fit is "weird context" work: problems that are hard because of the customer's environment, but still objectively verifiable. Codebase consistency across an organization is a prerequisite.

My Take

Factory's FDE motion is really about turning customer context into verifiable systems. The interesting point is not just that AI can do more work autonomously. It is that autonomy only works when the goal state can be tested.

That makes verification design part of the FDE job. The engineer has to understand the customer environment, define what "done" means, and create the checks that let inference run until the task is actually complete.