AI Shared Services

🚀 AI in Finance Recap: From Big Bets to Real Business Impact Over 300 colleagues from the magenta universe tuned in to this week’s AI in Finance stream, showing how much interest there is for AI-driven transformation in finance. The pace at which AI is improving is tremendous and it is accelerating exponentially. What is clear: Things will not stay as they are and there are massive opportunities for the teams that are able to adapt. Our event focused on how our products lay the foundation for being successful in this new and rapidly changing environment. 🔮 Delphi: Forecasting that proves its value Delphi showed that AI-powered forecasting works at scale. In several projects, including Devices, FTE and Cash Flow, Delphi already performed better than manual forecasts. What makes Delphi strong: • Swift delivery enabled by >20 reusable services • Delphi Studio: Our common frontend for standardized delivery • Forecasts increasingly piloted as a first opinion in planning processes And we are not stopping here: New use cases such as even more detailed scenario analysis together with our partners from T-Systems are coming up. 📊 Guided Reporting: Turning fragmented data into trusted insights Many departments sit on highly valuable data and process knowledge. However, much of this data is spread across Excel files, separate systems and manual handovers. Guided Reporting changes this by: • Structuring and connecting data sources to build common data models • Reducing manual handovers and repeatable work • Using AI and ML for forecasts, outliers and forward-looking insights The goal is simple: better decisions, less manual effort and solutions that fit into our existing tool and process landscape. 💬 ApolloGPT: Changing the way we act with Telekom Finance data The askT rollout showed the power of connecting apps like Sharepoint or Outlook to chatbots. But in daily business, we rely on many more Telekom Finance databases that are not yet directly connected. This is exactly where ApolloGPT comes in. We are building agents that: • Connect to Telekom systems and databases • Bring trusted, up-to-date data into askT, ChatGPT or other chatbot interfaces • Orchestrate information from different systems into one reliable answer 🤖 Agentic AI: From individual tools to orchestrated workflows We have built a strong foundation with our products. Now, we are moving from individual tools to agentic workflows across systems, data, decisions and actions. One example from the session: Vendor creation in SAP. A process with previously 52 steps was transformed into one agent-led process. This demonstrates the change enabled by Agentic AI: • From manual execution to orchestration • From switching between systems to agents working together • From human as executor to human as supervisor Want to take part in this journey and see how our products evolve? Then make sure to block your calendar: Our in-person AI in Finance event is coming up on November 4th. 📅⏰

🚀 Join Us: AI in Finance – Entering the Age of Agentic AI The next chapter of AI in Finance is unfolding — and you’re invited to be part of it. On May 21st, we are hosting a virtual AI in Finance session (1 hour), where our leadership and experts will provide an update on our AI in Finance Big Bets and take a deep dive into how Agentic Systems are becoming the next evolution step for AI in Finance. 💡 Why this matters AI in Finance is rapidly evolving from isolated use cases into a scalable and connected AI ecosystem that fundamentally changes how we work with data, workflows, and decisions. During the session, we will review the progress achieved so far across our strategic AI products and share our ambitions for the next phase of AI transformation. This includes: Updates on Delphi Forecasting, including forecast quality improvements achieved in 2025 and new use cases such as device forecasting ApolloGPT adoption and usage updates, with focus on Group Controlling and Segment Germany (TDG) Insights into how we plan to further integrate our AI solutions into broader DT initiatives such as askT A deep dive into Agentic AI workflows and connected AI experiences ⚙️ What to expect in the session In this webinar, we will share: Concrete insights into the evolution of our AI in Finance initiatives The role of platforms, integration, and our ecosystem approach How AI is evolving from standalone tools into connected workflows and intelligent systems What the next steps look like — from today’s capabilities to tomorrow’s operating model 🤝 Why you should join This session is designed for everyone interested in: The future of AI in Finance Practical applications of Agentic AI in enterprise environments How AI is evolving from tools to true digital co-workers How connected AI platforms can simplify and improve daily work Whether you are already working with AI or just starting to explore its potential: this is your opportunity to gain first-hand insights and engage with the community. 👉 Save the date: May 21st 👉 Duration: 1 hour | Virtual session 👉 Sign up via YAM: https://lnkd.in/d-qW3SXF We’re looking forward to seeing you there 🤩

🤝🚤 Building team spirit and stronger connections at AIS AI Shared Services is coming back energized from its recent team meeting in Brno 🇨🇿! Whether during informal exchanges, a shared team evening (with Karaoke 👀) or our joint activities, we had many opportunities to connect beyond daily work, strengthen team dynamics and collaborate across locations. ☀️ A special highlight was our team event • A boat tour, creating space for great conversations in a relaxed setting • A scavenger hunt, bringing together teamwork, creativity, and a bit of competition. 🧭 Clear direction from leadership A dedicated session with our leadership team provided valuable insights into: • How we will work together as one team • The organizational setup going forward • Our strategic priorities for the next phase ⚙️ Driving our products forward In multiple working sessions, we collaborated intensively on our products and use cases to refine concepts, align on next steps, and accelerate delivery. The in-person format enabled lively discussions and great brainstorming. ✨ Looking ahead We return from Brno with stronger connections, clearer alignment, and renewed momentum to continue building impactful AI solutions together. The next events are coming soon 🤩

🚀 From fragmented Excels to ML-powered rolling forecasts: Transforming people planning at scale People planning has long been shaped by complex, decentralized Excel landscapes. Different regions, different logics, limited comparability making it difficult to steer performance in time. That’s exactly the challenge we set out to solve at AI Shared Services. 💡 The starting point • Highly manual and Excel-driven • Fragmented across regions with no standard methodology • Limited to static forecast cycles (FC0 / FC4) • Lacking continuous visibility for timely decision-making ⚙️ What we built We introduced a centralized Rolling Forecast (RFC) solution: A unified platform that fundamentally changes how planning works: At its core, the solution combines machine learning-driven rolling forecasts that provide continuous, real-time visibility throughout the year, with a flexible React-based input interface embedded in Control Tower for intuitive user interaction. At the same time, it ensures data consistency through harmonization of multiple sources into one unified data layer, complemented by a built-in validation framework that guarantees high data quality across all inputs. 📈 Business impact • From static to continuous — forecasts available at any time • Standardized processes across all regions • Increased transparency enabling faster, data-driven decisions • Significant time savings through reduced manual effort • Stronger data quality through harmonization and validation   This is a clear step forward: Moving from manual, fragmented planning towards scalable, intelligent, and standardized forecasting. 🚀 Big shoutout to everyone involved: Sebastian Zeiss, Hannah Dickescheid, Max Ickenroth, Stefan Schulte and Jonathan Liese, Joaquín Soro Castelló, Laur Mircea Tudorascu, Dmytro Poslavskyi, Udo Taddigs, Sven Hüschemenger

Great to see AI taking center stage at our Annual General Meeting. Proud that our AI Shared Services team contributed today powering automated Q&A during the event — a real example of AI delivering value in a live, high-stakes setting. Thanks to the team and partners who made this possible: Hannes Wittig Philipp Schindera Dominik Sechser Fabian Stadler Udo Taddigs Sven Hüschemenger Luca Weigl Stanislav Zámečník Ioan-Tudor-Alexandru Anghel Monika Reháková Excited to keep bringing AI into real business use cases. #AI #Telekom #GenAI #Innovation

🚀 AIS Recap from the DTSE Townhall: Leading the Way into the Era of Agentic AI At this week's DTSE Townhall Udo Taddigs and Tobias Daum presented an exciting outlook on what lies ahead in the world of Agentic AI - and how we are actively turning this vision into reality. It was a moment that clearly showed: The next chapter of AI at Deutsche Telekom has already begun. 💡 But what is Agentic AI? Agentic AI represents the evolution from passive tools to active systems. Instead of just responding to prompts, AI agents can plan, decide, and execute tasks autonomously, working alongside us, orchestrating workflows, and unlocking entirely new levels of efficiency and innovation. 🤝 At the core of our approach is one key principle: human-centered AI. We strongly believe that the most impactful solutions are built with a human in the loop. Agentic systems are not here to replace people. Instead, they are here to empower us, augment decision-making, and free up time for higher-value work. 🛠 From vision to execution AI Shared Services has been driving AI innovation since 2019. Over the years, we have built strong foundations across use cases, platforms, and capabilities. Now, we are taking the next step: Working towards Agentic AI through our strategic Orchestra initiative and bringing together capabilities and expertise into a cohesive ecosystem that enables an enterprise-grade Agentic AI platform at scale. ✨ What’s next? If you’re curious to dive deeper into Agentic AI and see how it is transforming Finance and beyond, stay tuned: Join us at our upcoming AI in Finance virtual event this May. More details will follow soon - definitely worth keeping an eye on 👀 Let’s continue building, collaborating, and shaping the future of AI together 🚀 #AISharedServices #DTSE #AgenticAI #AI #Innovation #DigitalTransformation #AIinFinance

🚀 Another Milestone Achieved: AI-Powered Bad Debt Forecasting Goes Live in Poland! 🇵🇱 Following successful launches in Germany, Austria, and Croatia, we are thrilled to announce that our AI Bad Debt Prediction tool has now been officially implemented for T-Mobile Polska. This marks another significant milestone in our mission to minimize risk, sharpen forecasting precision, and empower finance departments across the Deutsche Telekom Group. 📊 This solution was developed by AI Shared Services in close partnership with our finance colleagues at T-Mobile Poland. By integrating internal KPIs with external data sources like Eurostat and Google Trends, the tool can project bad debt up to 12 months in advance—providing strategic insights that were previously unreachable through manual forecasting. What began as a response to the economic uncertainties of 2022 has evolved into a robust tool now optimizing operations in multiple countries. In Poland specifically, the rollout is already reducing manual workloads while significantly increasing predictive accuracy. 👏 A massive thank you to our colleagues at T-Mobile Poland for spearheading this successful rollout: Andrzej Reduta, Artur Bartoszewski and Marek Stusiński. Kudos to our exceptional cross-functional DTSE project team: Driven by Anna Sauer & Luca Weigl, alongside Catalina Maria Popa, Marcel Rešovský, George Rabat, Prajnya Nayak Maller, Daniel Kokoruďa, and Ladislav Farkas. Your dedication is truly shaping the future of finance. ✅ As more countries prepare for local implementation, this project continues to scale—demonstrating that AI is not just a buzzword, but a tangible driver of financial transformation. Also, a big thanks to our partner and supporters: Juraj Andras, Sven Hüschemenger, Udo Taddigs, Daniel Daub, Yoana Angelova, Thomas Geduldig, Fernando Calçada, Ivan Mijatovic, Renata Jelovecki Thuerriegl. 🔍 Stay tuned for more updates – and if you're passionate about the future of finance & AI, make sure to follow our journey here at AIS! #ArtificialIntelligence, #FinanceTransformation #BadDebtPrediction #Telekom #DTSE #AIinFinance #DigitalTransformation #Forecasting

AI Shared Services @ Inside Out at DTSE (Even on Rosenmontag 🎭) While others celebrate Rosenmontag, we are building the next chapter of enterprise-grade Agentic AI 🚀 Because AI at scale is not a side topic anymore — it is a strategic capability. At our “Inside Out DTSE Services” session, we are presenting our AI Shared Services Unit — and despite the parallel Carnival festivities, around 800 participants are joining 👏 🔥 800 colleagues. On Rosenmontag. 🔥 That is a strong signal for the relevance of what we are building together. We are no longer only talking about isolated AI use cases. We are shaping the move towards Agentic AI, scalable enterprise platforms, and the industrialization of intelligence across Deutsche Telekom Services Europe SE⚙️ As AI Shared Services, our ambition is clear: 🔄 Turning AI potential into scalable enterprise platforms 🏗 Moving from experimentation to industrialization 🤖 Building the foundation for Agentic AI architectures 📈 Delivering measurable business impact at scale A key strategic priority on this journey is close collaboration with other AI units across Deutsche Telekom 🤝 By combining strengths, aligning architectures, and sharing capabilities, we aim to create the best possible and most seamless user experience for our internal and external customers. Proud of what we are building — and even more excited about what comes next ✨ A special thank you to Luca Weigl for presenting our AI Shared Services journey and bringing the vision to life for such a large audience 🙌and to Alina-Iris Olf and Mirnes Dervisevic for organizing today’s session — a great platform to share what we are building together. #AI #SharedServices #AgenticAI #DTSE #DigitalTransformation

🤝 Human–Machine Collaboration in Practice: See what happens when practice and academia partner up 💡 For two years, the University of Münster and AI Shared Services partnered up to study how humans and machines work together - and what that means for the future of work 🔍 Research Focus How does human–ML collaboration unfold in practice, and what are its consequences for work? 👥 Who Was Involved The project brought together academic and industry expertise through the collaboration of Miriam Herding and Junior Professor Benedikt Berger with Telekom colleagues Udo Taddigs, Katja Steinert, Ruben Weinstock, Andrea Aßmus, and Yoana Angelova. Their insights were enriched by 35 interviewees spanning data science, project management, and financial controlling. 🧪 What We Did We took a deep look at financial forecasting as ML enters traditional processes. The study combined machine learning models with hands-on planning routines, revealing how people adjust, challenge, and enhance algorithmic predictions. Close collaboration with practitioners ensured that the research captured real decisions, real challenges, and real opportunities. 💡 The Insights we generated The team uncovered how ML tools can fit into established workflows, how trust and expertise evolve when humans work alongside algorithms, and how collaboration shapes outcomes - not just technically, but organizationally. 📍 Final Milestone The project wrapped up with an engaging presentation at the Telekom office, where the team shared key insights and opened discussions on what effective human–AI collaboration can look like in the years ahead. This research highlights a central truth: the future of forecasting isn’t just about smarter machines—it’s about smarter collaboration. 🚀

🎄✨ Wrapping up an incredible year at AI Shared Services ✨🎄 As the year comes to a close, it’s the perfect moment to pause, reflect, and say thank you🙏 Looking back on a successful 2025, we can be proud of the many achievements we reached together – impactful collaborations, innovative solutions, inspiring events and countless moments where ideas turned into real value. 🚀 What made all of this possible? A strong community, great partners, and teams who don’t just talk about innovation, but actively build it. 💡🤝 And this is only the beginning. We are convinced that 2026 will be even better – with new ambitions, stronger partnerships, and exciting opportunities ahead. We very much look forward to continuing our journey together with our partners and shaping what’s next. 🌍✨ But before we jump into the new year, there’s one important reminder: Take time to recharge, enjoy precious moments with family and friends, and slow down for a bit. The best ideas often come after a good break. 🎁❤️ Wishing you all a Merry Christmas, relaxing holidays, and a great start into the new year. Here’s to an even brighter 2026! 🎆 #AISharedServices #AIS #YearInReview #DTSE #ThankYou #MerryChristmas