You have an idea and need a working prototype
You want to show the product to investors, customers or your board. We build a version that can be tested with real users — without spending the budget on features that can wait.
Example:AgroSwarmTechnology partner for companies and startups
You have the idea and the deadline.What is missing is the team to build it.
We take the project from idea to working product — where software, AI and devices have to work together. Diagnosis in 1–2 weeks, a prototype ready for user testing in 6–12 weeks. One person is accountable for all of it, from the first conversation to deployment.
- 25 years
- at the intersection of electronics, physics and software
- 5 patents
- and patent applications in the founder’s name
- DESY · CERN
- control systems for the FLASH and European XFEL lasers
Speakersi — a language-learning platform with tutors and AI
Why clients come to us
We help when an idea needs to become a product
Most projects start from one of these four situations.
You run an R&D project and need a technology partner
You have funding and a schedule, and you need a team to turn the assumptions in the proposal into a working prototype. We understand how R&D projects and their milestones work.
Example:EcoCPRYour product is an app plus a device
Sensors, cameras, drones, lidars, edge AI. We select the components, connect hardware with software and take responsibility for the whole thing working in the field, not just on a desk.
Example:Sensor and compute stack for an autonomous robotic platformYou want to put AI to work in your company
Assistants for teams, document and data analysis, automation of repetitive work. We start from the process and a measurable outcome, and only then choose the model.
Example:AI assistant for recruitersProjects
From idea to a working version
Selected products we built from the concept up — in agriculture, medicine and aviation.
EdTechIn productionOwn product
Speakersi — a language-learning platform with tutors and AI
Video-lesson booking, tutor payroll, AI agents and an exercise library in one system
13 modules in one system — from CRM and scheduling to tutor payroll and reporting
AI & softwareIn productionWith a partner company
Lark AI EDU — training bots inside e-learning courses
Employees rehearse difficult conversations with a voice bot; managers get a scored report instead of a quiz
Each conversation scored against 5 rules on a 0–20 scale with written justification, computed once at the end
AgriTechIn productionOwn product
AgroSwarm
A drone-spraying marketplace — farmers book a treatment the way they book a ride
7 roles in one system — farmer, worker, operator, pilot, agronomist, advisor, administrator
HealthTechIn productionWith a partner company
EcoCPR
An eco-friendly CPR training manikin with an app that measures compression quality
The manikin withstood over 31,000 compressions by 400 people; mechanical tests point to over 100,000
Drones & aviationPre-launchClient project
SwarmATC
A drone flight-notification system connecting pilots with air traffic control towers
Rollout planned in South Africa, Zimbabwe and neighbouring countries
- LiDAR 3D
- Radar 4D
- GMSL2
- GNSS RTK
Autonomous systemsIn progressClient project
Sensor and compute stack for an autonomous robotic platform
3D lidars, imaging radar, GMSL2 cameras, GNSS RTK and edge AI — a complete perception and navigation stack for an autonomous vehicle
Multimodal perception — 3D lidars, 4D imaging radar and GMSL2 cameras feeding sensor fusion
How we work
Three stages. At each one you know what you get.
Every stage ends with a concrete result, so after each one you can decide whether to continue.
- 011–2 weeks
Discovery & concept
Conversations with you and future users, a workshop, a review of solutions and technical risks. We agree on what really has to work at launch.
Result
A document with architecture, scope, schedule and cost estimate
- 026–12 weeks
Prototype / MVP
We build in short iterations and regularly show a working version. Changes of direction surface early, while they are still cheap.
Result
A working prototype ready for user testing, with technical documentation
- 03depends on scope
Launch & growth
Production launch, monitoring, fixes and new versions — or a clean hand-over to your team.
Result
A product in users’ hands and a plan for its further development
The first call is free. The diagnosis is paid — but its cost is credited against the build if we do it together. You pay once, not twice for the same work.
No budget? Let’s talk about working for a share of the profits — we take part of the risk and scope the product so it starts earning. We do this selectively, for projects where we can see the market.
How we build
The speed of AI, the discipline of engineering
AI lets a small team work like a large one. The risk is that unreviewed code appears just as fast — so here, verification is part of the work, not a stage at the end.
AI agents work in parallel
Research, code, tests and documentation happen at the same time instead of one after another. That is why a working prototype takes weeks, without a department-sized team.
We verify content, not status
A green build and a 200 response are not proof. After every release we compare what is actually on the server with what we built, and run tests in a real browser.
Knowledge stays in the repository
Decisions, procedures and known pitfalls are written down next to the code. The project can be taken over without us — that is your safeguard, not our secret.
What you get
What we bring to your company
We do not sell developer hours. We take responsibility for building the thing, making it work for your users and keeping it maintainable.
A decision before you spend the budget
We start with a diagnosis: what is feasible, where the real technical risk sits and what to build first. You leave with a defined scope, a schedule and a list of things not worth doing.
A working prototype, not a slide deck
You get a version you can show to investors, customers and your board — and test with real users. First market feedback in weeks, not quarters.
Skills that are not worth hiring in-house
AI, computer vision, sensors and lidars, embedded systems, cloud. One team instead of four hires and three subcontractors blaming each other.
An R&D project that can be settled
We know how applications, milestones and settlements work — from both sides of the table, because we also review projects. Your grant schedule is our deadline.
Why Inn4U
You pay for the solution, not for our learning curve
These products, prototypes and patents were built earlier, in other projects. What you buy is the shortcut: proven patterns, partners already in place and decisions taken once — before they become your cost.
- digital products
- 30+
- At Open-RnD, including Booksy and iTaxi. The CarSync product moved to Arynga and was sold to Intel.
- and scaling
- Deployment
- We do not stop at the prototype: we go live, maintain and extend the system as users, devices or markets grow.
- patents & applications
- 5
- Computer vision, medical devices and AI-based animal monitoring.
- science & industry
- DESY · CERN
- Control systems for the FLASH and European XFEL lasers, NCBR expert, IT Centre director at Lodz University of Technology.
Who does the work
One person is accountable. One person does not do it alone.
You talk and work with the same person from the first conversation to deployment. Day to day, projects are run by a permanent three-person team. Partner companies join for larger builds, specialists I have worked with before take specific roles, and when a topic needs domain knowledge I draw on a network built over many years.
Partner companies
Experts and specialists
Top 500 Innovators
An association of scientists and innovation managers — experts across many fields of science and industry
EIT Health
Mentoring medical startups in the InnoStars and Starship programmes
NCBR · Innovation Coach
Experience as an expert and coach in programmes supporting R&D projects
Lodz University of Technology
Collaboration with the university, its specialists and student research clubs
Graduates in industry
Around 300 engineers I supervised during their studies now work at technology companies in Poland and abroad — that is where project teams and domain advice come from
Additional services
Expert knowledge beyond projects
AI training
Workshops and courses on the practical use of AI — for teams, senior management and universities.
Technology audits
Assessment of architecture, solution quality and technical risks of an existing system or a planned project.
Expert consulting
Assessing an idea and its potential, choosing the technology, planning a research and development project.
The first conversation
The first conversation already shapes the project
We don’t start with a quote — we start by understanding the problem. After the conversation you usually know more than before, even if we don’t end up working together.
The call is free — 30 minutes online, no preparation and no sales deck. Afterwards you get a written summary, and the decision to work together is not part of the call. If you prefer, we sign an NDA before we talk.
- 1
You understand the problem better
We ask about users, constraints and what has to work from day one. The idea takes a concrete shape.
- 2
You learn what not to build
A well-framed problem is often simpler and cheaper than the original concept. We say plainly what can be postponed or dropped from the scope — before costly decisions are made.
- 3
You get expert guidance
We point to possible technologies, risks and funding sources — and, when the topic needs domain knowledge, to experts from our network.
Let’s talk about your project
Tell us what you want to build. After the conversation you will understand the problem, the possible solutions and the next steps better.