Devot

Equestrian Show Manager (ESM) is a startup focused on building the world's best platform for digitizing and managing horse shows. The platform enables managing tournament arenas and stable/campsite registrations for a particular show. It helps show owners maximize their profits and provide the best experience for exhibitors, trainers, riders, and owners.

Besides building this complex platform from scratch, we resolved specific problems that occurred, like poor internet connectivity, because horse shows sometimes take place in remote and isolated locations.

Trainline sells train tickets and provides live train times and railway information through the website and mobile app.

They noticed a significant loss of potential ticket purchases because many users try to book their next ride while riding through remote areas with weak internet coverage. Users churn due to long loading time, so we detected the bottlenecks and built a special booking form for weak-internet situations. It behaved exactly like the existing site but had an adjusted framework and library to load faster. The page load time was reduced to only 0.84 sec on the 3G network (12.7 times faster than before).

Telemedicine platform that empowers virtual care services

We have helped the client gather and grow 7 independent teams in 15 months. The teams consist of over 90 experts covering everything from Ruby on Rails, ReactJS, iOS, Android, SCALA, project management and business analytics.

We take pride in the biggest Ruby on Rails team and the most extensive Rails project that has ever been developed in Croatia.

Software that enables clients to create their own forms which can later be sent to patients. Those forms are used for scheduling appointments or delivering additional documentation to medical doctors and institutions.

Collected data is processed and delivered to health facilities. It can be combined with other systems from the Rectangle Health portfolio, such as the payment of the patient management bridge.

Analyzing how people ask and await a response instead of searching for answers within existing documentation led to the chatbot named “Ante”.

Ensuring that the chatbot “Ante” could accurately identify and provide relevant answers required teaching AI about internal documentation and turning that documentation into embeddings so that the AI could understand and compare questions effectively.