Find out what your alarms miss.
A motor that vibrates a little more each week, still under its alarm limit. A freezer that starts warming at 2 a.m. Alerts that take too long to investigate. We work with your team on one problem like these, using your existing data or sensors we install, and test whether AI can show what's happening and what to do about it.
It's read-only, and your team stays in control.
Your team decides. Nothing changes on the equipment.
What would you like to understand better?
You may already have readings and alarms, but still spend time working out what they mean and what to do next. Or you may need sensors before you can see what is happening.
We start with one problem your team wants to investigate and a result we can measure.
An example: a machine drifting toward a stop
A motor on a packaging line starts vibrating more than usual. It's still under its alarm limit, so nobody is told. If it keeps going, it could stop the line at the worst moment.
With sensors on the motor, or the ones you already have, we could compare it with its own normal and look for what changed. Your team would get a likely cause and a suggested check, as in the diagram below. They review it, decide what to do, and record what they found.
It would test whether this gives your maintenance team earlier warning without more false alarms, and whether recurring cases can become rules your team trusts. We've tested this kind of fault in a simulated pump station.
This is a possible pilot, not a customer result.
An example: a freezer warming overnight
Since October 2026, attackers have remotely hit refrigeration in supermarkets across Europe, including hundreds in France. Fridges have been heated to as much as 50 °C, and stores have had to throw food away. Tamoz wouldn't stop an attack like this, because it isn't a security product.
A pilot would test a narrower question: with separate sensors in the cases, how soon could your team learn that something is wrong, and what would help them act? It could look like the illustration at the top of this page.
This is a possible pilot, not a customer result.
What Tamoz brings to the pilot
An alarm tells you a reading crossed a line. A rule catches a pattern someone already knew about. Tamoz is the AI agent we're building for what's left: it investigates what's likely happening, says how sure it is, suggests what to check, and learns from what worked. The decision stays with your team, because it can be wrong.
Your alarms and rules stay as they are. When the same recommendation keeps proving right, we'd review with your team whether it should become a rule: something they can read, question, and approve. That way, what your team learns stays with your plant. During the pilot, no rule would act on your equipment.
One motor, from a small change to your team's decision
Scroll sideways to see the full flow.
What we would do together
- Pick one problem. A line that keeps stopping, say, or alerts that eat up your team's time. Before we start, we agree what would count as success.
- Get the data. We start from the readings and records you already have. Where something's missing, we install sensors, following your site's rules.
- Watch and investigate. We collect a baseline, then test alerts and recommendations, read-only. Your team tells us which ones actually helped.
- Review it together. Compare the results with what you agreed, identify which recurring problems could become rules, and decide whether there's a useful next step.
What you'd come away with
- Recorded findings, and recommendations with their evidence
- A list of recurring problems that could become rules
- A clear review of what worked and what needs more investigation
What we'd need from you
- One person who knows the machine or process
- Access to your existing readings, or a place to install sensors
- A regular review with your team
Any later equipment action would need a separate agreement. And if the pilot doesn't help, you'll know that too, with nothing changed in how your equipment runs.
Who you'd work with
Three co-founders, working full-time on Tamoz, would work directly with your team on the problem, the software, and any sensors. We agree on scope and responsibilities before starting.
Nizar Zaidan: electrical and industrial systems. Electrical design for industrial and building projects, microcontroller-based controls, and measurement systems. LinkedIn
Ghassan Alhamoud: software and AI. Builds the Tamoz software and its simulator. Background in distributed systems and software architecture, with experience at Vinted and Wayfair. More about Ghassan
Alper Öztürk: growth and marketing. Founder of uberSEM, with an engineering and research background and experience in product marketing, lead generation, and growth marketing. LinkedIn
Why we're looking for a first partner
Tamoz has been tested in simulation and on a small hardware bench. We are seeking our first operational pilot and have not yet established savings in a real operation. A first partner helps us find out, with the founders' direct attention on your problem. The project overview explains the system and the evidence.
Tell us what you run and what you'd like to understand better. We'll talk through the problem and your data, and tell you honestly whether a pilot makes sense.
Discuss your operation