Hackathon AI: Maximizing Multi-Cloud Data Centers

Overview: We participated in the AI + Multi-Cloud Hackathon to explore innovative ways to maximize the value of Itaú’s CTMM data centers. With 20% of current energy capacity in use, there is room to expand and monetize through AI-driven products and colocation services. Our goal was to create solutions that generate revenue, improve operational efficiency, and position Itaú as a leader in AI and data sovereignty in Brazil.

Isabela Pascoal: PM – Agile, Data, AI, Business

Ste D’Elboux: Product Designer UX/UI – Design Thinking, Figma, Prototyping, AI

Ricardo Almeida: Product Designer UX/UI – Wireframing, Customer Journey, AI

Guilherme Monteiro: Data & AI – Machine Learning, Cloud, Python/R/SQL

Djairo Dantas & João Victor Caris: Fullstack Development

Problem: Underutilized data center capacity, high GPU costs, lack of market-ready tech product experience, and an evolving AI landscape.

Cloud Save Guard: AI-powered multi-cloud backup with predictive failure detection and voice-controlled automation.

AutoLab Cloud: Automatic creation of test environments with GenAI code review and versioned reset.

Downtime Detector AI: Multi-cloud resilience and self-healing deployments with natural language alerts via Slack/WhatsApp.

Optimized CTMM utilization and energy efficiency

Monetization via GPU-as-a-Service and colocation for third parties

Enhanced AI product offerings for finance and healthcare

Reduced operational costs and improved automation

Technologies: AI, Large Language Models (LLMs), Multi-Cloud (AWS, GCP, Oracle Cloud), Terraform, GitHub Actions

Outcome: Delivered a working MVP, live demo with synthetic client data, and actionable insights for Itaú’s CTMM expansion and AI strategy.