Delto and Microsoft brought together innovation leaders from Argentina's top banks in Buenos Aires to show transactional generative AI in production: LLMs, conversational AI Agents and a live demo where a skill is built in minutes. The takeaway: generative AI is already reshaping banking today, not tomorrow.
The transformation of the banking industry is no longer a future ideal: it is a reality happening now. At Delto we experienced it firsthand at our most recent event “Transactional generative AI in banking: future or present?” , which we held together with Microsoft, our technology partner, in Buenos Aires alongside innovation leaders from Argentina's leading financial institutions. In a close and exclusive format, we started the morning with a breakfast to discuss what is already happening in banks that dare to explore the frontier of generative artificial intelligence: from deploying conversational AI Agents to the real impact of LLMs (Large Language Models) on efficiency, scalability and customer experience. A tour of what is changing in the banking industry During the presentation, we covered key topics to understand why AI in banking is no longer just about efficiency, but a direct path to doubling innovation and profitability. The agenda included: Large Language Models (LLMs) We explained how these models learn through pre-training , what tokens are, and how neural networks and transformers operate. We also analyzed the differences between base models and assistants, which are fundamental to defining a solid technology strategy. AI Agents for banks We talked about what an intelligent agent is, how it directly impacts banks' results, and what concrete benefits —and the risks of not implementing them— are at stake. We also explored UX trends in conversational channels and how some institutions are already achieving visible improvements with generative AI solutions in production. Key challenges: trust, scalability and security We discussed how to address these axes through Azure-based architectures, modular skills and supervised models, following a scalable agentic technology approach. How does it work in practice? One of the highlights was the interactive live demo of the banking AI Agents suite, where attendees were able to see how to: Try out different skills working in a conversational environment. See how a new skill is created in minutes and tested instantly. Understand how these solutions integrate with existing banking systems in a secure and scalable way. These kinds of hands-on experiences make it clear why the next frontier in banking is agentic AI and how to move from concept to execution quickly. What's next? After the Q&A and closing remarks, it was clear that innovation and technology teams are ready to advance toward more mature, user-centered and truly differentiating solutions. In every conversation, every shared case and every demo, we saw that generative AI is already transforming the present of banking. At Delto we continue to support institutions that want to adopt this technology strategically, securely and with real impact. See you at the next event in Lima, Peru. In the meantime, we keep working so that generative AI continues to improve customer experience and transform the way customers interact with banks. Related reading: the regulatory map for conversational AI in LATAM · the AI agent platform for banking · adopt this technology strategically and securely .
What was the Delto and Microsoft event in Buenos Aires? It was "Transactional generative AI in banking: future or present?", an exclusive gathering Delto held with its technology partner Microsoft and innovation leaders from Argentina's leading banks. It featured sessions on LLMs, banking AI Agents and a live demo of the suite.
How is generative AI changing banking in Argentina? Banks are moving from pilots to production: they deploy conversational AI Agents and LLMs to improve efficiency, scalability and customer experience. The live demo showed how to build a skill in minutes and test it instantly, integrating it securely with existing banking systems.
How are the security and scalability of these AI Agents addressed? Through Azure-based architectures, modular skills and supervised models, under an agentic technology approach. This design resolves the trust, scalability and security concerns of banking technology teams.
What does the BCRA require from banks deploying generative AI in Argentina? Since 2023, Argentina's Central Bank enforces Communication A 7724, which requires banks to document every AI use, in house or third party, run impact assessments and adopt the three lines of defense model. Choosing a platform with compliance built in from the design stage speeds up approval by risk and legal committees.