Arman Karégar was taking his 12-year-old mixed breed dog to the vet in Stockholm when he had an idea. He noticed that he had to keep repeating basic medical information during the visits and wondered if any of this information was truly documented. It seemed like veterinarians spent excessive time taking notes but still managed to miss critical details about his pet.
So Karégar, a seasoned technologist, teamed up with college friend Max Xie to build Vetnio, a tool that records conversations between veterinary professionals and pet owners and generates clinical notes for the meetings. In one practice, Vetnio saved veterinarians approximately 40% of their time.
Currently, they operate across the UK, Germany, and several other European countries, and are quickly expanding in the US. On top of building their AI clinical notetaker, they wanted to build an assistant that could help with common veterinary administrative tasks. But this involved needing to build not just one but several agents — a booking agent, a reminder agent, a clinical-info agent. Each agent also needed to maintain its own context and tools, like a calendar API and medical database.
In February 2025, as they were building out this functionality, they turned to Mastra.
“We liked how Mastra handled complex workflows,” Karégar says. “We needed a flexible, extensible foundation for a veterinary-focused chatbot, and Mastra’s modular agents and workflow engine fit that need perfectly.”
How Mastra helps Vetnio go above and beyond transcription
Now a team of 9, Vetnio currently utilizes Mastra to deploy chatbots which assist with common veterinary administrative tasks, like scheduling appointments, booking follow-ups, sending appointment reminders, and answering common client inquiries.
The team was able to use Mastra to spin up specialized agents for each type of task and iterate quickly based on vet feedback. It’s often difficult to build a single agent that specializes in multiple types of tasks but they’ve found that Mastra makes it easy for them to build multi-agent systems.
Here’s a high-level breakdown of how they built the chatbot:
Frontend: Vetnio uses assistant-ui, an agentic chat frontend with a strong Mastra integration. Assistant-UI is responsive, so vets and patients can access the chatbot on both desktop and mobile.
Orchestration: Vetnio has set up a routing system where a Mastra workflow orchestrator routes each user request to the appropriate agent (booking, reminder, clinical info, etc). Each agent maintains its own context and tools, like a calendar API or a medical database.
Hosting: The Vetnio team is all-in on the serverless model, which ensures auto-scaling, low maintenance, and predictable costs.
“There were very few obstacles,” says founding engineer Emil Franzell. “Mastra’s documentation and examples covered most of what we needed, so setup and integration were straightforward. Agents and workflows could be added or modified without major rewrites, which kept development smooth.”
The team initially planned to build the chatbot from scratch. But after finding Mastra, making the switch was an obvious choice.
At the end of the day, the team is happy that Mastra helps them focus on what’s most important to them: building the best product experience for veterinary practices.