Grade Engine · sample mode · live scan rolling out Evaluation-led · Web · App · Web3

The work itself, labeled exactly as it is.

You won't find a borrowed client logo or a padded number on this page. You'll find a live product you can use today, a hand-built interactive demo, an engine still in build, and lab concepts marked as concepts. A studio that grades other people's work has to be straight about its own.

2 live · 1 in build · 2 concepts Status verified 2026-07-02

W-01 · Evaluation platform · Live

The Grader

An audit instrument with six lenses, one grade and a fix list ranked by what each fix is worth. Live on this site in sample mode.

Open the project file
Challenge
Development firms guess at quality, and buyers have no instrument to check the claim. So the job goes to whoever pitches best.
Approach
Six lenses and 26 weighted metrics, each carrying a confidence level. Every number is labeled Measured or Estimated. An estimate is never dressed up as a verified result.
Outcome
Live on houtoslabs.com. It runs in sample mode today, so you get a labeled preview rather than a measurement of your site. The live-scan engine (Lighthouse + axe-core) is rolling out.
System type
Evaluation platform · scoring engine · report generator · plain-language synthesis

W-02 · Analysis engine · In build

Peregrine Precision Evaluator

The measurement engine going in under The Grader. Instruments replace estimates, one stage at a time.

Open the project file
Challenge
Preview-mode grading has a ceiling. Without live scans, some readings have to stay Estimated, and the only structural fix is a real measurement backend.
Approach
One universal module contract feeds five analysis engines: accessibility (WCAG), privacy, EU AI Act, security and repository quality. Lighthouse and axe-core do the measuring.
Potential
Moves readings from Estimated to Measured wherever an instrument can take them. Parts of Reputation can't be measured off a site, and those stay labeled as estimates. Same rubric, tighter ranges on what each fix is worth.
System type
Analysis engine · crawler · module pipeline · AI synthesis

W-03 · Agent architecture · Concept — internal R&D

Autonomous Intelligence Desk

How far an autonomous system can be trusted, and what governance makes that trust rational.

Open the project file
Question
What would an autonomous system have to prove to earn each level of authority, instead of being handed it?
Approach
A promotion ladder: research, then simulated operation, then reconciliation, then limited live authority. Each step up requires measured performance. Enthusiasm doesn't count.
Status
Internal R&D. What we learn here feeds our client work on agent governance and evaluation suites.
System type
Agent architecture · risk governance · telemetry & reconciliation

W-04 · Memory architecture · Concept — internal R&D

Companion Systems

AI companions worth talking to, whose memory stays fully visible to the person they serve.

Open the project file
Question
Can a companion hold someone's interest while its memory and personalization stay fully visible to the user?
Approach
A hybrid memory architecture and consent-first personalization, with the safety layers designed before the engagement mechanics.
Status
Internal R&D. What we learn here feeds our client work on AI product design and responsible personalization.
System type
Memory architecture · personalization · safety design

W-05 · Interactive demo · Live

The Hidden Lab

One hand-written file holding a CSS-3D phone, canvas dashboards and a label on every simulated number.

Open the project file
Purpose
Proof you can poke at. The front-end and systems range you'd be hiring, in working form.
Approach
Canvas 2D charts, a CSS-3D device and streaming chat, all drawn in code. The labels come along too, with Simulated and Paper-mode marked throughout.
Outcome
Live on this site. One file, a 60fps target, and it respects reduced motion. The text still reads with JavaScript off.
System type
Interactive demo · canvas engineering · motion system
Enter the hidden lab Tip: there's a secret inside. ↑↑↓↓←→←→BA

Your system could be next on this page.

Bring us the ambitious one: the AI product, the protocol, the platform. We'll tell you what it takes to hold up once it's measured.