Julien HovanAI engineer · Montreal

Julien Hovan

I design the context, tools, boundaries, checks, and handoffs that let AI do useful multi-step work without making people lose control.

Harness / exploded assemblyAgentic engineering ↗Explore a part · follow a connection
contextWriting ↗
the few rules and sources that
matter for this step
permissions
the boundaries an agent
cannot cross on its own
toolsProjects ↗
the actions an agent can take
in the world
memoryBooks ↗
what stays useful
after a step is finished
checks
evidence that the work did the useful
thing, not just something convincing
loopNewsletter ↗
the next action, the result,
and the decision to continue
model
Tap a part to see what it does.

harness The workbench around an agent: tools,
boundaries, memory, checks, and decision points.