idea → development → success (short claim)
A year passed. Even so, it can take a long time to find the right partners for your project. Take, for instance, the company Perfect Canteen, which conceived the idea of an innovative smart refrigerator for dispensing boxed meals.
At Perfect Canteen, you can purchase lunch worry-free. You simply enter, scan your access, grab your lunch, and depart. The EatSmart refrigerator automatically identifies the items taken and forwards the information. It fell upon us to devise the entire technical infrastructure, from integrating all the hardware to developing the mobile application.
If it weren’t for Cookielab, we might have scrapped the project long ago. But Radek, Kolda and their team have partnered up with us to push the product forward. They don’t shy away from challenges and they understand that cooperation isn’t only about development, but also about things like presenting to investors. In hindsight, I’m really glad I trusted the recommendations and my impression from the first meeting. The professionalism, insight and friendly approach of the people at Cookielab is exactly what a person is looking for in their developers.
Eat Smart
In the IT world, Perfect Canteen is best known for the canteen it operated in Avast (now Gen). The founder, Filip Sajler, is quite well known from the cooking show Kluci v akci. However, Filip did not want to stay at the canteen. He thought: “What if businesses could use a smart fridge to find great restaurant-quality food?” without canteens, and the worry of payment -- I just come, take my food, leave. This idea meant developing a globally unique solution, because classic vending machines did not make economic or operational sense.
Using a template is meaningless when dealing with a completely new product. Therefore, we first threw ourselves into the so-called Product Discovery phase, during which we define the problem we are trying to solve, priorities for development, and a roadmap for the future of the product. In this phase we also validate the technical feasibility.
With EAT SMART, we needed to present a working technology (proof of concept) to management within three months. The alpha and omega was to create a system that allowed for the purchase to take place as quickly as possible and without the need for a payment terminal. We assembled a wooden prototype and set up communication between individual IoT elements while our development team focused on developing the application.
During storage, the refrigerator automatically recognizes the stored food thanks to the RFID tags placed on the boxes. At the same time, this method speeds up the service as much as possible – refilling takes only a minute and a half.
When picking up food, the employee just needs to place their employee card next to the refrigerator and it will automatically log them in. Alternatively, they can log in through the application by scanning the QR code. Users view the current menu, filter foods according to allergens, the refrigerator lights up where the selected food is located, and then it’s time to eat.
The refrigerator automatically locks and sends information about the food taken to the employee and to the database.
The personnel department receives instructions on how much to deduct from the employee's wages, and the supply department receives a message on how many meals are left. This whole process takes a few seconds, the slowest purchases take 15 seconds. No payment terminals are needed, and no food code punching in. Everything happens automatically.
The EAT SMART fridge also collects all gastro-specific data – monitoring the temperature and how long the door has been open. Everything is further evaluated so that people can be sure that the food is still in perfect condition.
After the initial prototyping, we were convinced that we had solved the problem of reliable chip reading, but the technology used did not work reliably in a real environment. Specifically, we had to solve a situation where we could not read all the RFID chips quickly and reliably enough, and the position detection within the refrigerator did not work at all.
We were surprised by how big and fundamental an influence the physical properties of the materials used would have on us. The metal elements reflected and attenuated our signal. The black paint used in the interior and on the boxes contained graphite, which is electrically conductive, which also affected the signal. In addition to all of this, the adhesives used caused us a problem. Suddenly we were at a dead end. There were two options: either we shut down the whole project, or we go back to researching other available technologies.
Curiosity won, and since there are no coincidences, the client came across a video by German RFID expert Olaf Wilmsmeier, who presented a new type of LOCFIELD® RFID UHF antennas. We reached out to him and after trying it, which ultimately led us to our "Eureka!" moment.
So, we started from scratch. We set up the integration of individual technologies and continued to develop the mobile application. We tested everything constantly to make sure that there would be no dramatic complications once we put the refrigerator on the HOPI Holding campus for our pilot operation.
In the pilot operation, we tested how accurately the refrigerator could recognize the food taken. Here, it turned out to be essential that those who produce and pack the food also cooperate in some way on this project. The success of automatic recognition also lies in a correctly affixed RFID tag. It must not only agree with the food (so that there is no pork dumpling on the box with tenderloin), it also needs to be positioned correctly so that the antenna can see it. For convenience, we have added the option to pay via Apple Pay and Google Pay.
By April 2025, 100 EAT SMART refrigerators should be distributed in the Czech Republic. Perfect Canteen's vision that every company will be able to afford to provide its people with restaurant-quality food, even if they are in an industrial hall or need food 24/7, is taking shape. It's up to us to tame the hardware and connect everything so you don't have to worry about a thing when buying food. In short, you come, beep, take your food and leave. A simple idea, not quite a simple solution. But we like such projects. 🙂
All startups are striving for the fastest go to market possible. Carvago, an online marketplace with used cars, was no exception.
When the company was struggling to find developers for their product, our team jumped in to provide knowledge, expertise, and processes. With the need to build on code from the previous agency — not from scratch as we’re accustomed to — it was not an easy task. But that’s what we like.
“Our choice fell on TypeScript, which enables us to safely rewrite larger parts of the application. We added runtime validation responses from API, and introduced industry standard libraries. The client got a modern React with functional components and hooks, and we reduced the giant root state for faster and safer development of new features. With regular automatic app updates, we gained access to new features speeding up development and performance for higher traffic.”
Co-Founder, Cookielab and CTO as a Service at Carvago
The client came to us without any documentation and with unfinished source codes and infrastructure (FE - Next.js & React.js, BE - PHP & Symfony 2, infra - Terraform 0.10). Our team needed to analyze quickly what was already done to develop further.
With infrastructure, we had no other option than to deploy & fail & repeat until success. The challenge were the processes or rather the lack of them. No tools, no processes, no monitoring was in place. Time to start building them as the first releasecommit was due in just two weeks!
After analysis, we needed to set proper procedures for development. Our DevOps team got the infrastructure up and running, but mid-process we found out that a large part of the services were missing in the Terraform and we had to add them. Next, came product management, in which we:
collected requirements from key stakeholders
validated them against the actual state of the application and its ecosystem
proposed priorities for development.
Next, we set up the development process with SCRUM methodology and Jira workflows to be able to scale the team. Then came the transfer from BitBucket to GitLab, and from ECS to Kubernetes — both of which improved the overall developer experience. Kubernetes was especially helpful for easier implementation of Review Apps, and better tooling for management, deployment and monitoring.
After putting out fires, we finally compiled the technological roadmap to stabilize development and deliver new features predictably.
This case study details just the beginning of our cooperation with Carvago. Stay tuned for more details on web improvements - e.g. getting infrastructure ready for influx of visitors during nationwide campaigns, and user testing.
