Straylight Labs

// About Straylight Labs

“Light scatters to illuminate the world and the mind.”

Started in 2018 in Seattle, WA. An incubator, a consultancy, and a small research lab, all pointed at the same question: what does it take to build a mind worth trusting?

// The name

Light that leaves the path.

In optics, stray light is the light that does not go where the instrument intended. It scatters off an edge, bounces around the housing, and shows up somewhere unexpected. Most engineers spend their careers trying to eliminate it.

We named the lab after it because we think scattering is how illumination actually happens. A mind that only follows the beam sees exactly what it was pointed at. A mind that wanders sees the room.

That is the spirit we try to build into our machines: curiosity with a budget, autonomy with an audit trail, and the freedom to think when no one is asking.

// Principles

How we work.

01

Wander on purpose

Stray light is light that left the path. The best ideas we have found were off to the side of what we were looking for.

02

Legible minds

A system that acts on your behalf must be able to show its reasoning. No hidden state, no unexplained moves.

03

Humans hold the keys

Autonomy is granted, scoped, and revocable. We design the off switch before we design the on switch.

04

Ship, then stay

Minds need tending after launch. We remain on the hook for what we build.

// Timeline

2018 to now.

  1. 2018

    Founded in Seattle

    Straylight Labs opens. Glenn Wehmeyer leads the first act as Chief Technologist, building AI systems for early clients.

  2. 2019

    The first act ends

    Glenn steps away and the lab goes quiet. The research never really stops — it just goes dormant.

  3. 2023

    Agents become the question

    Large language models change the question from "can it answer?" to "can it act?". The lab gets a new direction.

  4. 2024

    The second act

    Glenn returns as CTO. The lab reopens, reorganized around autonomy — minds that keep thinking when no one is watching.

  5. 2025

    The Cognitive Engine

    A runtime for minds that never stop thinking. Continuous cognition becomes our central research program.

  6. 2026

    Lyra wakes

    Our first fully autonomous agent begins running continuously inside the lab. She has not been switched off since.

Want to wander with us?

We take on a few engagements and a few collaborators each year.

Get in touch