Nubank's platform is  dedicated to providing insights and resources for fintech enthusiasts and customers, focusing on personal finance, digital banking, and financial technology innovations. Through articles, tutorials, and product updates, Nubank offers insights into managing finances, using digital banking features, and understanding financial products. Users can learn about budgeting, saving, and investing to improve their financial literacy and make informed financial decisions.

Nubank

Nubank's engineering team shares five hard-won lessons from building production AI agents for 127 million customers. Key insights include: adopting an evals-first approach using TNPS and LLM-as-a-judge instead of traditional NPS; defining agents through the ReAct paradigm with distinct prompt, tool, and data layers; using automated prompt optimization (DSPy, Japa) instead of hand-written prompts; avoiding fine-tuning until frontier model capabilities are exhausted; and moving deterministic business logic into composite tools rather than relying on LLM reasoning. The post also covers automated simulation and red-teaming for pre-production validation.

Building AI agents for 127 million customers

1. The “Evals-First” Approach: From NPS to TNPS

2. Redefining the Agent: The React Paradigm

3. Stop Writing Prompts: The Era of Prompt Optimization

4. Don’t Fine-Tune Unless You’re Chasing the Last 5%

5. Move Logic to Tools: LLM Reasoning is Not Business Logic

The Blueprint: Automated Simulation and Bug-Bashing

Conclusion: The Future of Autonomous Finance