Enterprise AI,
from R&D to production.
We build AI systems for the R&D teams of large technology companies: agent platforms, protocols, inference infrastructure. From the first experiment to a system that runs in production every day.
Worked with
Production Notes
What we learned building AI systems in production, by the engineers who built them.
Latest note
From faster agents to productive teams
The work did not vanish when agents took it on. It moved.
11 min read
Read the note
SW factories
Writing code by hand is already obsolete. Next comes the unattended era, where code ships to production without a human reviewing it at all.
Radek Ježek & Matouš Havlena · 11 min read
The research loop is going AI-native
AI is moving from summarizing papers to running the experimental loop itself. What researchers need to confidently delegate more of an investigation.
7 min read
Selected work
Agent frameworks, robotics and real-time systems, built inside R&D teams.
Agents
Agent Framework
Orchestrates autonomous agents across multi-step tasks, with governance and full observability.
See the projectRobotics
Robotics Arms Orchestration
Real-time choreography of 15 robot-mounted 4K displays.
See the projectReal-time
Processing Video and Sensor Data
Live video and sensor telemetry in one view for autonomous chemistry labs.
See the projectWho we are
A startup inside enterprise R&D: senior, hands-on, in a hurry to learn what is true.
- 01
Senior and hands-on
Engineers who have run AI systems in production inside large organizations.
- 02
Startup pace, enterprise gates
We onboard ourselves, pass security review, and ship without slowing down.
- 03
Substance over hype
Weak bets die in writing. What we learn goes into the Production Notes.
Let’s talk.
Tell us where it gets hard. You will talk to engineers, not a sales team.