Title: Hrvatski Telekom: What a simple fleet management chatbot can teach telcos about selling

URL: https://www.infobip.com/blog/hrvatski-telekom

Every day, a small team at Hrvatski Telekom's (HT) car fleet department handled a steady stream of phone calls from colleagues. The questions were always the same: when do I need to change my tires, where do I take my company car for a service appointment, and so on. For a telco that champions digital automation, it was exactly the kind of problem worth tackling.

Magdalena Sekulić Ljubić, Product Owner in HT's ICT division, has spent nearly five years building tools that close these gaps. The company car management chatbot is the latest example of what it looks like when a telco uses what it sells.

## Innovation that starts in-house

For most organizations, digital transformation is something they do for customers. For a telecommunications company, the expectation runs both ways.

HT's ICT team has been working on that expectation for four years. The approach is straightforward: product owners run regular internal pitches, walking colleagues through the tools they manage and looking for problems those tools could solve. The goal is to turn deep knowledge of Infobip's platform into concrete improvements for HT's own operations, so they can confidently recommend those same solutions to customers who face similar problems.

## Addressing the repetitive issues first

The fleet management chatbot idea started as a casual conversation between co-workers.

During one of HT's internal pitches, a colleague from Sekulić's ICT team ended up talking to someone from the car fleet department. The fleet manager mentioned, almost in passing, that his team spent a significant portion of every day on the phone answering the same questions, over and over.

It was exactly the kind of repetitive, high-volume workload a rule-based chatbot is built to handle. This wasn't HT's first time building an internal chatbot. Sekulić's team had already deployed some for ICT services like FAQs and billing queries. The pattern was familiar: a team drowning in repetitive tasks, a clear set of questions, and a channel employees already used every day.

WhatsApp was the natural fit. Most HT employees were already on it, and the team had enough internal deployments behind them to know it worked.

## Built quickly with zero coding

The car fleet team sketched out the chatbot flow themselves. Sekulić's ICT team gave guidance, then used Infobip's chatbot building platform to turn that sketch into a working tool.

It took the team less than a month from their initial conversation to launching the chatbot. The chatbot is intentionally simple. An employee saves a WhatsApp number, types “hello,” enters their registration plate, and gets back the exact information they need like tire change location, or service appointment details, without waiting for someone to pick up a phone.

The fleet team was enthusiastic from the start. They weren't being handed a system from above; they designed the flow, understood the logic, and owned the result.

## Employee reception and adoption of new automation

HT launched the chatbot with a single announcement on the company's internal pages. No campaign. No training sessions. Within days, several hundred employees had already used it.

Three months after the initial launch, the chatbot logged around 600 sessions with a 75% engagement rate.

What continues to surprise her is how little effort it took to get there. Although they've had success with other automated chatbots on WhatsApp, it's always surprising to see how a simple chatbot can alleviate workloads and make the employee experience significantly better.

## Why you should practice what you preach

The chatbot does something beyond taking over repetitive tasks from the fleet team. It gives HT's customer conversations a different kind of credibility.

When HT recommends Infobip's chatbot building platform to enterprise customers, it does so as an organization that has built, deployed, and measured the tool internally. This is what four years of internal chatbot development has produced: a team that can point to its own operations as the proof of concept.

On AI, Sekulić is clear-eyed about when it makes sense, and it's not always the most advanced solution. The goal of the fleet chatbot is to handle a defined set of questions, and a rule-based flow is exactly right for that. But as HT's internal automation ambitions grow, she sees the potential behind AI-powered chatbot for more complex use cases.

For now, her advice to other organizations considering the same path is simple:

## You might also like...