Built around how you make things.
Running inside your walls.
Scheduling software for job shops. Knowledge tools and configurators for engineered-product makers. Your data never reaches a public AI model.
Illustrative example
What we build
One way of building, two kinds of manufacturer.
Product for job shops
Cadence
Advanced planning and scheduling. It builds a finite-capacity schedule from your ERP, re-plans when something slips, and checks every move against your constraints. Your planner decides.
How Cadence worksEngagement for engineered-product makers
Company brain
A private assistant built on your specs, procedures, past quotes and the people who've been there longest. It answers with citations, and says when it doesn't know.
What a company brain doesEngagement for engineered-product makers
Configurators and calculators
The rules, tables and engineering checks behind your products, built into the way you already design and quote. Deterministic code, signed off by your engineers.
How we build configuratorsMost AI is built to sound right. Ours is built to be checked.
How we work
AI reads and explains. It never does the arithmetic.
The same five rules apply to everything we build, whether it's a schedule, an answer or a calculation.
Private by design.
Models run on hardware or a private cloud you approve. Your data doesn't go to a public AI service, for any of our products.
The math is checked by math.
Schedules are checked by rules and a constraint solver. Configurators and calculators are deterministic code. AI reads documents, finds sources and explains results, and that's where it stops.
A person says yes.
Nothing changes in your ERP, your schedule or a design until a named person approves it. We log who approved what, and when.
Every answer shows its evidence.
You see the sources, how fresh they are, and which checks ran. When the evidence isn't there, the system says "insufficient evidence" and asks someone to fill the gap.
One product line, fitted to you.
We configure software we maintain instead of writing a new codebase for every customer. So fixes and updates reach you without a rebuild.
Product first. Services that make the product better.
Every engagement fits one of three shapes, and we tell you which before any work starts.
Included
Configuration
Most of what we deliver. Your routings, calendars, priorities, approval rules, report formats and units, set up in software we already maintain.
Priced as implementation
A reusable extension
When you need something we haven't built, we build it as a capability the next customer can use too. That keeps it maintained long after your project ends.
Quoted separately
One-off work
Work only one customer will ever need. We quote it on its own and tell you plainly what it costs to support. We turn down work that would split the product in two.
Your data stays where you put it.
We win by being useful, not by making it painful to leave.
You own it
Your data, and anything we derive from it, belongs to you. Export it whenever you like.
Nothing shared by default
We don't train anything shared on your data. Aggregate learning happens only with separate, written permission, and it's never a condition of service.
No public AI APIs
Inference runs inside the boundary you approve. Our engineers build and test with synthetic data, and they don't get access to your production records by default.
Commercial work only
We don't accept ITAR-controlled technical data, and we screen other export restrictions before anything is loaded.
Straight answers.
Both, on purpose. We build software we maintain and configure it to each customer. When something new is needed, we build it so the next customer can use it too. One-off work is priced separately, and we say no to work that would fork the product.
On hardware or a private cloud you approve. We agree one supported deployment pattern with you before quoting, including who owns the hardware, upgrades and support.
Not by default. Using your information for anything beyond your own deployment needs separate written permission, and refusing it doesn't change your service.
No. Everything we build reads from the systems you already run. Cadence starts read-only and doesn't write back.
We serve commercial manufacturing. We don't accept ITAR-controlled technical data, and we screen other restrictions before any data is loaded.
With a paid, bounded scope: one problem, one data route, success measures we agree once a baseline exists, and a decision at the end to continue, narrow or stop.
About Quogen
Quogen is a small, self-funded company building AI systems for manufacturers. We're not chasing every corner of manufacturing software: no ERP replacement, no marketplace, no promise to run your whole operation from an app. What we do is narrower. We fit software to how a manufacturer actually works, keep it inside their walls, and make sure a person can check every answer it gives.
Tell us about the problem.
Talk to usor write to info@quogen.ai
Say whether it's about Cadence, a company brain or a configurator, and we'll take it from there.
We reply to every message ourselves. If we're not the right fit, we'll say so, and usually suggest who is.