X 000Y 000

// SUBJECT 001 DETECTED

CHIEF AI OFFICERORGANIA · WEON

STATUS: EXPERIMENTING●

POINTS0

MORPH000%

ALT024 M

LINHAS04
EXPERIMENTOS20
RODANDO13
25.4284° S · 049.2733° WBRT

CITY 1:500 // REBUILDING

HSLab · open AI lab

HélderSato.

Scroll to take apart

01 — The labLAB/2026

I test AI inthe real world.

And report what works, what breaks and how much value is left at the end. Here live the experiments, the math, and the questions about what this changes for us.

Research lines

Experiments

Articles & essays

Running now

01 / 04 — Research line · EH · Running

Hybrid teams.

What has to exist around a model so humans and agents can work on the same team, with trust?

12 experiments

Open the lineOpen the line In practice: squadhibrido.com.br ↗

02 / 04 — Research line · AD · Validating

AI adoption.

How do you bring AI into a mid-size company, area by area, proving value before talking money?

5 experiments

Open the lineOpen the line In practice: organia.ai ↗

03 / 04 — Research line · IR · Designing

INOVA-R.

How do you turn fast AI innovation cycles into validated capabilities that enter existing products with provable value?

3 experiments

Open the lineOpen the line

04 / 04 — Research line · SI · Coming soon

AI Simulation.

Can we simulate customers, teams and markets with AI agents to decide better — without confusing simulation with prediction?

Coming soon

Open the lineOpen the line

01 / 04

02 — ExperimentsEXP · 020

On thebench.

All experimentsAll experiments

CodeExperimentStackStatusDate

EHHarness for hybrid teams12 · Running →EH-12The internal agent serving the websiteAG-UI · Widget · TurnstileRunning01.11.2026 EH-08AI risk, privacy and cost in one placeAI governance · LGPD · AuditoriaRunning31.10.2026 EH-01When an agent fails, the model is rarely the causeLangGraph · Middlewares · CIRunning16.10.2026

ADAI adoption method for business5 · Validating →AD-03A system that guides whoever applies the methodFastAPI · React · Postgres RLS · LangChainRunning24.10.2026 AD-04An agent that maps and redesigns processesAgentes · Redesenho · Humano no circuitoRunning20.10.2026 AD-05Simple AI governance for mid-size companiesGovernança · LGPD · AI literacyRunning16.10.2026

IRINOVA-R — innovation built into existing software3 · Designing →IR-01Moving a solution from organIA into WeONWork Pilot · WeON · PRTPlanned03.11.2026 IR-02What to hand over along with the codeSpec-Driven · ADR · TestesPlanned03.11.2026 IR-03Skills for agents to move a solution into another productClaude Code · AGENTS.md · SkillsPlanned03.11.2026

SIAI Simulation — simulate before decidingComing soon →

03 — Articles & essaysART · 000

Worth readingslowly.

All articlesAll articles

04 — RadarGITHUB

Taking apartopen source.

See the radarSee the radar

a2aproject/A2AStudyingA2A

Agent-to-agent protocol across systems. On the radar, not yet built into the platform.

EH · Harness for hybrid teams↗ag-ui-protocol/ag-uiUsingAG-UI

Agent–UI protocol. Powers the visual blocks and Live Apps in the OrganIA platform.

EH · Harness for hybrid teams↗langchain-ai/langgraphUsingLangGraph

Stateful agent orchestration. Part of the agent-builder execution core.

EH · Harness for hybrid teams↗

05 — Who writesSUJEITO 001

FIG. 01 — HOVER

Hands-onbefore opinions.

Chief AI Officer and co-founder of and . Building customer service software since 2004. MSc in Distributed Systems, PUC-PR.
500+ brands and 10k daily users on the platform. That is where experiments meet reality.
Research, prototype, validate with clients — only then it becomes product. This lab is the public side of that loop.

Off the lab: , , and whiskies. They are hidden around here — go look.

06 — StageKEYNOTE · WORKSHOP · FIRESIDE