Oracle Fusion Cloud

Oracle bets on AI integrated with data to accelerate adoption in enterprises. Event Data Deep Dive highlighted how infrastructure, security and governance became central pieces to scale AI projects and autonomous agents in companies. 05/22/2026 14:25 The race for artificial intelligence in companies is no longer just about generative models or new chatbots. During the Oracle Data Deep Dive, held in São Paulo last Thursday (21), Oracle presented the vision that AI, data, infrastructure and security have stopped functioning as separate layers to operate in an integrated manner through the same corporate architecture. "The problem is not AI, it's the foundation," said Alexandre Maioral, president of Oracle in Brazil, during the opening of the event. The speech summarizes the company's assessment that many companies accelerated AI initiatives without resolving structural bottlenecks, such as fragmented bases, difficulty in integrating systems, and lack of governance over corporate data. Throughout the meeting, executives and experts discussed how databases, cloud and autonomous agents are changing the way corporate applications are developed. In Oracle's assessment, organizations without an architecture prepared for AI will face difficulties in scaling projects, connecting internal information, and maintaining control over increasingly automated systems. AI closer to data. This view was deepened by Shasank Chavan, vice president of global database technologies at Oracle. According to him, the new generation of intelligent applications requires databases capable of understanding context, meaning and relationships between data. "The real impact happens when AI works with private data, keeping it protected," said the executive. The company's strategy is to bring artificial intelligence to the data - not the other way around. This means incorporating AI resources directly into the database, allowing intelligent agents to query information, automate processes and execute tasks faster and more securely. During the keynote, Chavan highlighted the use of AI vectors - structures capable of transforming documents, images and videos into representations understandable for artificial intelligence models. The proposal is to unify traditional and semantic searches in a single environment. According to Oracle, this model reduces the operational complexity created by the use of multiple specialized databases. "Data chaos slows innovation and hinders AI," he said. Another point raised was the need to build more reliable environments for autonomous agents. For the company, systems capable of making decisions and performing tasks independently require fast access to data, as well as security and governance mechanisms integrated into the architecture itself. "The database needs to understand meanings, not just patterns," the executive highlighted. Autonomous agents and new software engineering. The event also addressed the advancement of AI agents, systems capable of planning actions, making decisions and operating autonomously. For Oracle, this evolution should change business applications and the very logic of software development. During the presentation led by Leandro Vieira, senior vice president of AI and Tech Cloud for Latin America, and Marcelo Christianini, vice president of AI and cloud engineering for Latin America, the executives argued that the market is moving towards more conversational and automated interfaces, where users stop operating systems and delegate tasks directly to intelligent agents. "It may be that I just talk to the system, and the agent performs the actions automatically," said Vieira. The executives also discussed the difference between so-called "vibe coding" - development driven by speed, improvisation and experimentation - and "agentic engineering", an approach focused on structuring applications prepared to operate with AI agents long-term. According to Oracle, the two approaches can coexist, but require solid foundations to avoid fragile applications that are difficult to sustain as they scale. Among the pillars pointed out by the company for more robust AI projects are infrastructure, governance, system design and operational control. "AI can write the code, but it's these layers that determine whether it will survive," Christianini said. Security and quantum computing. Security appeared as another central axis of discussions. Oracle executives warned about the future impacts of quantum computing on current encryption systems. According to Shasank Chavan, the combination of quantum computing and AI agents could pose a significant risk to traditional digital security infrastructures. In response, Oracle said it is already incorporating quantum-resistant encryption into its databases. The company also highlighted cloud operation models that allow different levels of control by clients, from jointly managed environments to fully autonomous databases. Shared knowledge. In addition to the keynotes, the Oracle Data Deep Dive included practical workshops focused on testing Oracle AI Database features and discussions on corporate AI applications. The event also marked the launch of the Oracle Innovators Club, an initiative created to connect developers, experts, tech communities and companies in an ecosystem focused on practical learning, access to technology and creation of opportunities. According to Oracle, the proposal is to create a continuous environment for exchanging knowledge about AI, cloud and data infrastructure, bringing together from students and early career professionals to technical leaders, freelancers, entrepreneurs and more experienced market specialists. The initiative comes at a time when the rapid evolution of artificial intelligence has been pressing technology professionals to update technical skills and expand connections within the ecosystem. In this context, the Oracle Innovators Club seeks to function as a meeting point between different profiles of the tech community, encouraging collaboration, networking and development of projects related to the new generation of AI-based applications. Startups is part of the routine of over 30 thousand decision makers. Will you stay out?

Oracle Financial Services has launched new agentic AI capabilities for corporate banking, introducing pre-built AI agents for treasury, trade finance, credit and lending operations. The platform aims to automate mission-critical processes and accelerate decision-making for financial institutions and corporate banks. Key innovations include AI agents that extract and validate data from lengthy loan contracts, pull financial metrics from statements, monitor external news sources for risk signals, and generate credit memo narratives. For trade finance, agents validate bank guarantee applications and create supply chain finance programmes from sales contracts. The enhancements are part of Oracle's next-generation banking platform, which embeds AI directly into customer engagements and business processes whilst maintaining human oversight. Oracle plans to release hundreds of additional corporate and retail banking agents within the next 12 months.

Oracle has launched AI-enabled capabilities in its Primavera Unifier platform to help project and asset management teams improve compliance and efficiency in capital projects. The new features include AI-driven workflow summarisation, enhanced data integration, and improved safety monitoring. The updates enable teams to prioritise critical activities through AI-powered business process summaries that provide structured chronologies of project progress, including participant details, timestamps and decisions. Enhanced Oracle Integration capabilities allow real-time data automation across enterprise systems with audit trails. Additional features include no-code process design tools, dashboards for performance tracking, and safety monitoring through Oracle Construction and Engineering Advisor. The platform aims to support high-compliance projects in regulated environments such as federal programmes and aerospace.