ROI should not be the only determining factor for AI projects, says Pavel Jindra from CETIN

October 24, 2024

He understands that artificial intelligence is transforming the playing field, deciding the winners and losers in the race for technological dominance. For businesses looking to do more than just stay competitive—those aiming to lead by building a top-tier internal AI department—the path forward can seem daunting. How can companies prepare for this future?  

Pavel Jindra, Manager of IT Platforms at CETIN, a company focused on the operation and management of telecommunications networks in the Czech Republic and a key player in developing digital connectivity, offers a strategic look into how to harness AI for long-term success. 

Read the interview

(This text has been edited with the help of ChatGPT.) 

Ivana Karhanová: You mentioned in our pre-recording conversation that we now have a chance to introduce new technologies. Why now? 

Pavel Jindra: It's not necessarily about now, but as a technology company, we continuously strive to adopt new technologies to maintain our competitiveness and improve both internal processes and services. Specifically, regarding AI, the market is flooded with offers from various companies and consultancies, a trend driven by the rise of ChatGPT a few years ago. While the media buzz is significant, we're careful not to be swayed by it, but it's essential to keep an eye on the market and explore the possibilities. 

Ivana Karhanová: AI is a broad term. What exactly are you focusing on right now? What do you need to build? 

Pavel Jindra: As I mentioned, we are a technology company and the owner and operator of the largest telecommunications network in the Czech Republic. We have vast amounts of network data from signalling, which offers opportunities for machine learning. So, we're more focused on machine learning tasks and advanced analytics. We're also experimenting with more common tools like co-pilots, assistants, and chatbots, but that's not our core business. Since we don't have retail customers, we don't need a chatbot for daily customer interactions. 

Ivana Karhanová: Machine learning and optimization aren't new—these topics existed before ChatGPT's models were published almost two years ago, in the fall of 2022. Why are you implementing these use cases now? 

Pavel Jindra: I think the popularization and widespread adoption of AI among the general public helped, as it resonated at the business management level. It became a big topic. To simplify it, if you want to sell something, label it as AI, and success is almost guaranteed. It's similar to the "bio" label in food. This has helped us in a positive way because it's now getting attention. If we discuss this within the company, everyone understands that if we don't jump on the train now, we risk missing out, and it will be hard to catch up later. 

Ivana Karhanová: But your leadership understands the importance of AI and that CETIN should be active in this area, correct? 

Pavel Jindra: Absolutely. This isn't something we just started focusing on in the last few weeks. The preparation was longer, and the turning point was at the end of last year when we agreed to launch a comprehensive program covering not just AI but also digitalization, automation, and robotics. All of these activities were consolidated under one umbrella program. An action team was formed, comprising colleagues from IT platforms, enterprise architecture, and transformation. This sent a clear signal within the company that AI is a major priority. There's also a budget that supports us in implementing proof-of-concepts (PoCs) and adopting AI and other digitalization initiatives. Now it's up to us to deliver concrete results and justify this to management. 

Ivana Karhanová: Could you have done this without their support? 

Pavel Jindra: Definitely not. We wouldn't have the resources or capacity. It would be challenging to push these activities forward, and it probably wouldn't make sense to even attempt them. 

Ivana Karhanová: You mentioned that you want to build an internal AI team. Where are you starting? 

Pavel Jindra: We've already started. Initially, we said, "We have a vision, but let's break it down into concrete actions." The first step was to create an AI strategy, which is the foundation. We established this in the first quarter of this year, laying out a plan for the necessary skills, roles, and organizational structure to support AI efforts within CETIN. 

Ivana Karhanová: You developed this strategy with Adastra. How did that process go? 

Pavel Jindra: It began with several sessions and workshops involving both business and technical teams within CETIN to assess our starting point. This was crucial for tracking our progress over time. Working with Adastra, we outlined five key dimensions to cover, and the final output was a maturity assessment, showing where we stand and setting goals on a 1-5 scale. 

Ivana Karhanová: Are these goals tied to measurable KPIs? 

Pavel Jindra: Yes, there's a framework in place to help measure progress. Simplified, we're currently at level 2 on a scale of 1 to 5, and we aim to reach level 3 within a year. 

Ivana Karhanová: What do these levels represent? 

Pavel Jindra: They reflect our readiness in terms of data, technology platforms, and processes. It's a high-level view of how we intend to set up KPIs, monitoring, and platform operations, including governance, legislation, and security. Ethical considerations are also essential, and communication strategy plays a big role in sharing successes internally, which helps in gathering use cases. 

Ivana Karhanová: It sounds like a significant project. 

Pavel Jindra: It is. And this is just the foundation, setting the stage for where we are and where we want to go. Along with the AI strategy, we also developed a roadmap for the next year, focusing on adoption by May 2024. It covers setting up teams, governance, and technology, along with initial use cases. 

Ivana Karhanová: How many use cases have you gathered? 

Pavel Jindra: We've collected 13 use cases, which we've prioritized using the RICE matrix to assess costs, time, financial and non-financial benefits, and feasibility. 

Ivana Karhanová: How long did this process take? 

Pavel Jindra: About 6 to 8 weeks, including final reporting. 

Ivana Karhanová: Could you describe one of the use cases? 

Pavel Jindra: A simple example is a chatbot for our HR department, designed to assist new employees during onboarding. It answers common questions like how to request vacation or what benefits are available. We decided to use Copilot Studio, as it offers a quick, low-code solution. 

Ivana Karhanová: How do you measure ROI for these projects? 

Pavel Jindra: We look at it from two perspectives: business process improvements, which lead to cost savings or increased revenues, and network technology enhancements that improve service quality or minimize outages. However, ROI should only be a supporting argument, not the sole deciding factor. The long-term benefit of adopting new technologies is crucial. 

Ivana Karhanová: Has your management accepted these softer benefits? 

Pavel Jindra: Yes, even though they are less tangible, they've accepted them. Our goal is not only cost savings but also enabling employees to focus on more creative and meaningful work, which could lead to future innovations. 

Ivana Karhanová: What would you have done differently, looking back? 

Pavel Jindra: The main surprise was capacity issues. We had great plans but lacked people with the right skills. Many stakeholders weren't fully committed, which slowed progress. In hindsight, I'd focus more on securing full commitment before starting the projects. 

Ivana Karhanová: Thanks for the great conversation and see you again! 

Pavel Jindra: Thanks for having me, goodbye.