Sign In
Register

Partnership opportunities

Secure your seat

Call to action
Your text goes here. Insert your content, thoughts, or information in this space.
Button

Back to speakers

Alberto
Romero
Director, Generative AI Platform Engineering
Citi
Alberto has a comprehensive background in Data Engineering, Big Data, Machine Learning, and GenAI, with 15 years of experience building real-time platforms, with a focus on streaming analytics and machine learning applications including Fraud Detection, Risk Prediction and High Frequency Trading systems. Previously, Alberto co-founded an AI startup and served as its Chief Technology Officer (CTO) for 6 years, building from scratch an awarded insurtech platform that leverages data and AI to predict risk exposure of vehicles in real-time based on ML behavioural analytics and geospatial models. The startup was acquired by a large multinational at the end of last year. In January of this year, Alberto joined Citi as Director of GenAI Platform Engineering. At Citi, Alberto is responsible for designing and building GenAI platforms for the bank, leading initiatives in the space and developing and productionising AI products. His role also involves staying up to date about technological advancements in GenAI, evaluating new frameworks and techniques to address the needs of internal customers. Some of his work includes the development of AI agents, complex RAG pipelines and evaluation frameworks to measure success of AI applications in production-like environments.
Button
07 November 2024 10:00 - 10:30
Optimising GenAI outcomes in financial services with DSPy
As the financial services industry embraces GenAI to drive innovation and efficiency, the need for robust and trustworthy outcomes becomes paramount. Traditional RAG pipelines, while powerful, often fall short in addressing the unique challenges of the finance domain, such as data and model interpretability. DSPy emerges as a game-changer, offering a comprehensive solution tailored to the intricate requirements of domain-specific use cases. By seamlessly integrating LLMs with structured programming constructs, DSPy empowers developers to build GenAI applications that are not only highly performant but also more predictable, auditable, and compliant. This talk delves into the architectural advantages of DSPy over traditional RAG pipelines, highlighting its ability to optimise prompt instructions, enable fine-grained optimisation, and facilitate interpretable decision-making processes. Through real-world use cases and code examples, we will explore how DSPy's unique features, such as signature-based programming, chained LLM calls, and optimisation techniques, can unlock new frontiers in GenAI adoption across Financial Services. Attendees will gain insights into the transformative potential of DSPy in optimising GenAI outcomes, ensuring regulatory adherence, and fostering trust in AI-driven Financial Services.