Often it is something that already works, like a pricing routine from 2019 or an overnight job that grew from twenty minutes to four hours.
We deliver a tenfold return. Each euro you spend with us brings back ten, as an analysis, as software or as a whole system we run for you.
Don't take our word for the engineering. Read the code.
Where our software runs
100M+ people
use Livesport. We built nine of its internal systems. We have also built for Guaranteed Rate, Zimpler, Cloudpermit, Crossbeam and Stylitics.
How we check the maths
We wrote our own theorem prover.
It is called Rainbow. Every number we give you goes through it first.
See RainbowWhat we do
01 The Studio
Custom software for the part of your business that makes or loses money. We work out what a better way would earn you on your own figures. Then we build it and run it. It is most of what we do.
See what we solve
02 The Decision Lab
We try exact algorithms and machine-checked proofs on real company data. The ones that win go into the systems we build for clients.
Enter the Lab
What we solve
Thin margins and late projects usually start with a decision somebody makes by feel, every day.
Project margin
For each project, what it will cost to finish and the chance it ends in a loss. You see it while a manager can still act.
ReadCapacity and scheduling
What to accept, schedule or defer when you are short of people and time.
ReadRemediation and warranty
Defects, claims or cases, ranked by what each one costs you while it waits.
ReadCritical software
A machine-checked proof that the rule holds for every possible input.
Reade-Fuels and hydrogen
e-SAF, e-methanol and hydrogen plants, modelled against real weather and power prices, from design to daily dispatch.
ReadIf it is not on this list, tell a partner about it. The first conversation is free.
How we work
We find, prove and build. Then we operate what we built and keep making it better. Each stage has its own fee.
Two to three weeks for a fixed fee. We look at the operation and the money in it, and come back with a number and a plan.
Four to eight weeks. We test the riskiest part on your own data, with a solver, a prototype, a proof or a shadow run.
What worked becomes production software, with monitoring and a manual override built in.
Work
What we sell
Each product started as client work that paid off.
For founders and investors
Shows where your engineering money goes and what the business got back for it.
Explore GitrevioOpen-core theorem prover
A theorem prover where proofs are ordinary data. It comes with kernel-checked theorems about all 222 catalogued operations-research algorithms.
See RainbowWe build in Clojure. Rainbow, orhof and chachaml are open source, and the book behind them is Applied Higher Order Functions.
Why Flexiana
The partner who frames the problem stays on it through the build and after.
Your first payment buys an answer to the riskiest question. The big work comes later, if that answer says it is worth it.
The best projects start with something a company has done the same way for years, because it works well enough. So I ask what a better way would be worth.”
Jiri Knesl, founder of Flexiana
Author of Applied Higher Order Functions
FAQ
A consultancy gets paid for the report and an agency for the build. We get paid to make the operation earn more than it costs. And the people who price the work are the people who write the software.
The first conversation is free. The audit starts at €15,000, fixed before we start. In two to three weeks you get a map of the decision and what it costs you today, a ranked list of fixes, and our advice to go, reshape or stop. If you go ahead, the fee counts against the bigger engagement.
Look at what we have built. We wrote nine internal systems for Livesport, which more than 100 million people use. The hard parts of our own tools are open source, and you can read a worked sample of the audit findings. And the first step is a fixed-price audit, so you can try us on a small piece of work.
Good, we work best with one. We bring operations research, formal methods and product engineering, and we turn the analysis into a system that runs. Your team knows the business and keeps the baseline we build.
Where to start
A partner reads it and tells you what it would be worth to fix.
