Skip to content

Sixi Lab R&D · Locarno, Switzerland

AI agents your company can govern.

Sixi Lab R&D designs, tests and documents AI agent systems for regulated companies, and builds its own tools to do it. Hire the lab on contract, for a fixed-scope review or to build to order.

Radek BrusSenior Software Engineer & AI ArchitectOpen to contract and consulting work, remote across Switzerland and Europe.

Sixi AssureDemo design

Who is leaving?

Acme Consulting (vendor): 23 roles, 14 with nobody else named, bearing on 4 agents.

Owner ofAlso named
Agent registrynobody else named
Key vaultnobody else named
Model endpointnobody else named
Analyst specialistHead of Digital Operations

Agent passport
Calculation specialist

Owner
Core platform team
Credential
federated, lifetime 3600 s
Kill switch
not declared
Tool scopes
not declared
From the demo design of Sixi Assure, with made-up names: what a vendor still holds on the day it leaves, and what one agent's passport does not say.

Work with the lab

Three ways to work together

You work directly with the person who designs it and builds it.

  • Contract

    On your team

    A senior engineer and AI architect inside your team for a few months: architecture, hands-on build and reviews, on your stack and in your process.

  • Consulting

    A fixed-scope review

    One clear question, a fixed scope and a written result that your own people can act on.

  • Build to order

    Built for you

    You describe the outcome and the lab builds it: a prototype, an internal tool or an edge system, delivered with tests and documentation.

What the lab takes on

Governance, security and engineering for AI agents

  • AI governance and agent management

    An inventory of your AI agents: who owns each one, what it may touch and how it is stopped, with evidence that a risk or audit team can read.

    See Sixi Assure

  • Secure agent platform architecture

    Agent identities, scoped tools over MCP and A2A, governed gateways and evaluation before release, on Microsoft Foundry or Google ADK.

  • Security testing of AI agents

    Red-teaming over REST, MCP and A2A as a regression test in CI, judged from what went over the wire.

    sixi-scanner on GitHub (external site)

  • Calibrated evaluation

    Decisions with an honest confidence value instead of one model acting as judge: local encoders first, the cloud only for uncertain cases.

    laya-as-judge on GitHub (external site)

  • Local and edge AI

    Open-weight models on NVIDIA Jetson and DGX Spark, for data that must not leave the site.

    sixi-asi on GitHub (external site)

  • Off-grid radio and embedded systems

    LoRa mesh radio, signed messages and firmware for field devices: the work behind Rifugio and Faro.

    About Rifugio

Projects

What the lab builds

Products, field work and open-source tools, in the order we would show them to you.

  1. Sixi Psyche

    Coming soon

    A research project where AI meets psychology. More when it is ready.

  2. Sixi Assure

    Flagship

    Know who owns every AI agent before the people who built it leave. A model of your agent system at design time, deterministic rules with their clauses, a passport for every agent and a signed handover pack. Offered as a 2-week review and as a service.

  3. Rifugio

    Field project

    A way to stay in touch when the internet or the mobile network is down: signed TEST alerts and help messages over LoRa mesh radio, for municipalities and for families. In development.

  4. sixi-asi

    Open source

    Runs open-weight language models on local NVIDIA hardware (Jetson Thor, DGX Spark): a speculative-decoding inference server in C++ and CUDA.

  5. laya-as-judge

    Open source

    A decision model whose confidence you can check: calibrated evaluation and routing for typed decisions, with local encoders and selective cloud verification.

  6. sixi-scanner

    Open source

    Red-teaming of AI agents from one command-line binary: single-turn and multi-turn probes over REST, MCP and A2A, made for CI.

  7. And the rest

    More

    Faro, Sigillo, sixi-observer, the public map of mesh radio networks and the tools around them.

Who you work with

The person behind Sixi Lab R&D

Radek Brus
Senior Software Engineer & AI Architect

14 years of engineering, including 3 years owning end-to-end AI platforms in regulated sectors (medtech, industrial automation, building technology).

Builds AI systems for strict governance: identity and controlled gateways at the core, agent workflows on top, evaluations before release, compliance from day one.

Before Sixi Lab

  • Johnson Controls
    Cloud & AI Architect
  • Danone
    AI Solution Architect
  • Codit
    Consultant & Architect

Certified

  • Microsoft Multi-Agent AI Solutions Expert (AI-500)
  • Azure Solutions Architect Expert
  • Azure Security Engineer
  • AWS Security Specialty
The Madonna del Sasso sanctuary on its rock above Locarno, with Lago Maggiore and the mountains in a soft haze behind it.
Locarno on Lago Maggiore, where the lab is based.

Tell me what you are building.

We start with a call of 30 minutes: what the system does, what it touches and what could go wrong. Then you get a fixed-scope proposal.