Side-by-side comparison of how Brazil's three largest banks use artificial intelligence on WhatsApp to serve 74 million customers
26 de fevereiro de 2026
10 min read
Source: microsoft.com
74M
Bradesco BIA Clients
90%
Itaú Pix Recurrence
-70%
Nubank Response Time
The race for artificial intelligence in Brazil's banking sector has three clear protagonists: Bradesco, Itaú, and Nubank. Each adopted a distinct strategy for integrating AI into WhatsApp, and the results are impressive at different scales.
Bradesco, with BIA (Bradesco Artificial Intelligence), co-developed with Microsoft and Avanade, achieves 90% resolution without human intervention for 74 million customers. Itaú bet on Pix via WhatsApp with 90% recurrence and 150 GenAI solutions in production. Nubank, partnering with OpenAI, reduced response time by 70% and resolves 50% of Tier 1 queries without escalation.
74M
Bradesco BIA Clients
90%
Itaú Pix Recurrence
-70%
Nubank Response Time
Brazilian banks face a unique challenge: operating at continental scale with 24/7 service expectations. Bradesco, with 74 million customers, processed over 2 million daily requests. Itaú needed to make Pix transactions faster than its own app. Nubank, with explosive growth, handled 2 million chats and emails monthly.
The common denominator was the need to scale without compromising quality. Hiring proportionally to growth was financially unsustainable. AI was the only viable answer.
Bradesco — The Multi-Agent Platform: BIA uses Azure OpenAI Service with 10 deployed LLMs and over 400 experiments. The "Bridge" platform, co-developed with Microsoft and Avanade, operates as a multi-agent system. The bank is also piloting "Smart Pix" for voice-command Pix transfers on WhatsApp.
Itaú — Conversational Pix: Itaú created "Inteligência Itaú," a proprietary GenAI platform with multiple foundation models. WhatsApp Pix accepts voice, text, and images, plus bill splitting and instant receipts. With 1,800 AI models and ~500 data scientists, the bank has the largest AI operation among Brazilian banks.
Nubank — The OpenAI Partnership: Nubank uses GPT-4o and GPT-4o mini for chat and email support, RAG for enterprise search, copilot for call center agents, and GPT-4o vision for fraud detection. WhatsApp Pix was tested with 2 million users.
**Bradesco:** 90% resolution without human intervention, 95% accuracy, 8x growth in BIA usage after GenAI, 8x productivity increase for branch managers, -40% in service time (pilot), 2 billion tokens processed daily.
Itaú: WhatsApp Pix 20 seconds faster than the app, 90% recurrence, 97% accuracy in Investment Intelligence, 79% funnel conversion, +141% increase in GenAI volume.
Nubank: -70% response time, 50% of Tier 1 queries resolved without humans, 2.3x faster resolution, -60% in WhatsApp transaction processing time.
All three banks demonstrate that AI on WhatsApp is not an experiment — it's critical infrastructure for Brazil's financial sector.
Catalisa offers a platform that enables banks and fintechs to replicate this level of conversational AI sophistication without building from scratch. With AI Agents for service, transactions, and advisory, Workflows for multi-agent orchestration, and Building Blocks for core banking integration, the platform accelerates AI adoption in the financial sector.
Native WhatsApp Business API integration and multi-LLM support allow financial institutions of any size to implement intelligent agents with the same quality as major banks.
MICROSOFT. “Banco Bradesco SA — Azure AI Services”. microsoft.com. Disponível em: https://www.microsoft.com/en/customers/story/19177-banco-bradesco-sa-azure-ai-services. Acesso em: 27 fev. 2026.
TI INSIDE. “Itaú coloca 150 soluções de GenAI em produção”. tiinside.com.br. Disponível em: https://tiinside.com.br/12/01/2026/itau-coloca-150-solucoes-de-genai-em-producao-e-amplia-uso-de-ia-em-141/. Acesso em: 27 fev. 2026.
OPENAI. “Nubank Case Study”. openai.com. Disponível em: https://openai.com/index/nubank/. Acesso em: 27 fev. 2026.
Schedule a demo and see how Catalisa can transform your business
Quantitative analysis of the real return on AI investment in customer experience, with consolidated data from Forrester, Boticário, Lotte, and Synthesia.
27 de dez. de 2025
·
10 min read
How Synthesia, Lotte, and Carrefour absorbed exponential growth in service demands without proportional hires, using AI as a force multiplier.
04 de fev. de 2026
·
9 min read
Analysis of how AI agents on WhatsApp evolved from reactive bots to proactive business consultants, with cases from iFood (Gerente, Cris) and Magazine Luiza.
11 de jan. de 2026
·
9 min read
© 2026 Catalisa. Todos os direitos reservados.
Conforme LGPD
Meta WhatsApp Business
Dados no Brasil
We use cookies for analytics and load external resources (images) to improve your experience. By clicking "Accept", you consent to the use of cookies and third-party resources in accordance with our Privacy Policy.