DevAfterWork at Betfair Romania Development: Data, Infrastructure & Staying Human in the Age of AI

What happens when AI becomes part of the way we build products and manage infrastructure? 

On 15 October, we welcomed the local tech community to our office for another DevAfterWork session, organised together with DevTalks. Under the theme “Data, Infrastructure & Staying Human in the Age of AI”, the evening brought together different perspectives on a question increasingly relevant across technology: as AI becomes embedded in our workflows, how do we use it to work smarter while keeping human expertise, context and judgement at the centre? 

Four perspectives on working with AI 

Connecting the ideas throughout the evening was our colleague Anca Bradisteanu, Senior Engineering Manager, who moderated the event.  

After a welcome note from Radu Miclaus, General Manager, Anca kicked things off with an interactive quiz before guiding the audience through the four presentations. 

The evening featured four talks from Betfair Romania Development colleagues, looking at AI through the lenses of infrastructure, software engineering, data and product. 

Raluca Temle, Senior Infrastructure Product Owner, opened the talks with a session on how AI blends into to the infrastructure game. As teams move from provisioning CPU, memory and storage towards working with models, GPUs, tokens, context, data and intelligence, the shift also calls for a rethink of areas such as CI/CD, observability, scaling and platform engineering. 

Next, Daniel Andreca, Principal Software Engineer challenged us to think about how we can use AI to amplify engineering skills while continuing to exercise the thinking that makes those skills valuable. 

Maria Sulea, Senior Data Engineer, talked about a practical balance between automation and human decision-making. AI can take repetitive work off our hands, but important decisions still need human context and judgement.  

Finally, Razvan Man, Product Owner, brought the Product perspective and focused on a fundamental ingredient for making coding agents effective: context. He explored how Product can adapt its deliverables for agentic engineering workflows, turning product intent, technical constraints and design inputs into clear, purpose-built briefs. 

From presentations to conversation 

The talks were followed by a panel Q&A with all four speakers, creating space to dig deeper into the ideas presented during the evening and explore practical questions from the audience. 

And because DevAfterWork is as much about the community as it is about the content, the evening continued over snacks, drinks and networking, giving participants and speakers more time to exchange ideas and keep the conversations going. 

Keeping humans in the loop 

Across infrastructure, engineering, data and product, the evening offered different perspectives but returned to a common idea: getting value from AI isn't simply about doing more with new technology. It's also about being intentional about what we automate and what we delegate. 

A big thank you to DevTalks for partnering with us to bring another DevAfterWork to our office, to Raluca, Daniel, Maria and Razvan for sharing their experience and perspectives, and to Anca for guiding the conversation throughout the evening. 

And, of course, thank you to everyone who joined us. Your questions, curiosity and conversations are what turn an evening of presentations into a genuine community exchange. 

We look forward to welcoming you back at our office for more conversations, ideas and opportunities to learn from one another.