Highlights from Data Day Texas 2025

It was an incredible opportunity to connect with the global data community, exchange ideas, and explore how advanced data strategies can reshape the energy landscape. During my session, I explored how AI-driven probabilistic forecasting and optimisation enhance BESS trading. To do this, I showcased a real-world solution that integrates real-time and historical data with AI-powered forecasts, enabling traders to navigate complex energy markets and maximise revenue. The aim of this talk was to highlight the role of automation in energy decision-making, making storage asset optimisation more efficient and scalable.

Data Day Texas was packed with thought-provoking sessions from industry leaders, sparking discussions that will no doubt shape the future of data and AI. Here are some of my highlights from the event:

Key takeaways from Data Day Texas 2025

Automating Financial Reconciliation with Linear Programming and Optimization, Bethany Lyons

Bethany Lyons presented a session where she addressed the complexities organisations face in linking multiple payment requests to corresponding payments – a process often fraught with data quality challenges due to missing relationships. Bethany showcased how mathematical techniques from operations research can automate this reconciliation process, reducing manual effort and minimising errors. Her real-world examples showcased the practical applications of these methods across finance and financial services, highlighting a path toward more efficient and accurate financial operations.

Introduction to Data Contracts, Mark Freeman

Mark Freeman‘s keynote on data quality opened up exciting discussions about data contracts and the potential of data products in improving governance and scalability. He introduced the data contract architecture, guiding the audience through practical applications tailored to various data use cases, from startups to large enterprises. Mark shared real-world success stories of companies implementing data contracts in production, offering strategies to overcome common challenges and secure leadership buy-in for critical data quality initiatives. Attendees left with a clear understanding of how to define data contracts and integrate them into their data ecosystems to enhance governance and scalability.

Deployment at Scale of an AI System Based on Custom LLMs: Technical Challenges and Architecture, Arthur Delaitre

Arthur Delaitre provided an in-depth look into deploying AI systems at scale in his session. He discussed Mirakl’s approach to transforming seller catalogue onboarding by implementing a scalable AI system utilising custom fine-tuned Large Language Models (LLMs) and state-of-the-art multimodal models. Arthur explored key topics such as infrastructure deployment for scalable LLM inference, model fine-tuning to reduce hallucinations, and micro-service orchestration for efficient operations. His insights provided a comprehensive view of overcoming scaling challenges and ensuring high performance in real-world AI system deployments.

How to Become a Hero – The Journey to Text, Bill Inmon

Bill Inmon, the legendary “father of the data warehouse”, captivated the audience with his session, “How to Become a Hero – The Journey to Text.” He illuminated the often-overlooked value of text data, discussing how organisations can unlock significant opportunities by analysing unstructured text. Bill emphasised that while text requires its own technology and techniques, the rewards are substantial, including increased profits and customer acquisition. His talk served as a call to action for organisations to delve into text analytics and discover the hidden treasures within their unstructured data.

Interactive Town Hall, Joe Reis & Matthew Housley

Joe Reis & Matthew Housley set an engaging tone for the day with their interactive Town Hall session. They created a collaborative space where attendees could discuss emerging trends, challenges, and innovations in the data world. This interactive format encouraged open dialogue, allowing participants to share experiences, ask questions, and explore new ideas in data engineering and analytics. The session exemplified the spirit of community that defines Data Day Texas, leaving attendees raring to go for the day ahead.

Til next year, Data Day Texas

Data Day Texas 2025 reaffirmed the power of data to drive transformation across industries. As we continue to push the boundaries of data and AI in energy, we’re excited to integrate these insights into our work and contribute to the evolving conversation. A huge thanks to the speakers, organisers, and everyone who made the event a success. We can’t wait for next year!

Stay tuned for more updates from the Hypercube team as we continue to explore the intersection of data, AI, and the energy sector. To hear more about energy technology insights and events, join our Beyond Energy Community.