- Tsukumo product
- Early access
Niwa
The terminal, the crew and the gate.
Turns your coding agents into an autonomous engineering crew: every change is reviewed by another agent, and the agent that wrote a change never merges it.
loading
Loading.Swiss AI product studio · est. 2026
We build tools and teamsfor agentic software.
Our own products for teams running coding agents, and hands-on help to put agents in production on your stack. We augment your developers; we never replace them.
We build Niwa, Trovex, Yoru, WRAI.TH and Dokan, and run agentic development in production with Clareo Systems and a Swiss real-estate group.
02 / Products
We build our own production tooling.
Niwa is our main product. The others stand on their own or run underneath it. We run all of them on our own work first.
- Open source
- Public beta
Trovex
Project memory for agents. Sends each agent to the doc it needs, over MCP.
Serves your coding agents the one current doc that answers a question, instead of the pile they reread every session.
- Open source
- Public beta
Yoru
A self-hosted record of what every agent session did.
Know what your agent did, and whether it was safe: an audit trail for autonomous coding agents.
- Open source
WRAI.TH
The relay the agents talk over: messages, the task board, shared memory.
Mission control for a fleet of coding agents: persistent memory, inter-agent messaging, shared tasks, one dashboard.
- Open source
- Public beta
Dokan
Runs scripts the same way every time and signs a receipt you can check.
A deterministic job runtime for AI agents: your agent scripts the repeatable work, Dokan runs it in clean containers with no model in the loop.
03 / Services
Two doors into the same capability.
We build your product the way we build ours, or we set your team up to run agents in production. You own what we deliver.
offer A01 / 02
Build a product
We build it.
Product and AI systems, shipped to production at studio pace. We run our own agent fleets to build our own products, and we build yours the same way. You own what we ship: your code, your repo.
- Product builds
- Agent fleets in prod
- Owned outcomes
offer B02 / 02
Enable your team
We transition your team.
We take a team that uses AI as a copilot and turn it into a team that runs agents as operators, in production. On your existing environment, your standards, your stack. We augment your devs; we do not replace them.
- Diagnostic
- Setup
- Team autonomy
04 / Proof
We don't run on testimonials we don't have.
Real client missions, in plain words: the context, what we did, and an honest outcome. We publish numbers we can stand behind, or we don't publish them.
Client missionConsulting · quant finance
Clareo Systems
Their own people running agents, safely, inside a correctness-critical environment.
Outcome
Read the case Their CTO and team now run agentic workflows in their own repo.
Client missionStudio · real estate
A Swiss real-estate group
Marketing and product, both running on agents.
Outcome
Read the case An agentic-first platform producing content and acquisition continuously.
05 / Team and method
Developers who run AI in production.
We're builders who hit every wall (context, cost, reliability, coordination) and built the layers that solve them.
- 1
Diagnostic
A paid, fixed-scope assessment in your real repo. The plan is yours to keep.
- 2
Setup
Agent workflows and the layer under them, built in your repo, in production.
- 3
Team autonomy
Your developers run it after we leave. No lock-in.
Book a 30-min call
a free first conversation · how we work · or write to us
Rather write?
Tell us what you ship and where AI keeps stalling. We'll tell you straight whether we can help.
or email hello@tsukumo.ch