XConf Americas 2025

by technologists, for technologists

As AI reshapes the work of software teams and brings both opportunities and uncertainties, ensuring high-quality delivery becomes even more essential. This session shows how to adopt AI-first practices while keeping developers in control and aligned with the engineering principles that support effective delivery. Using real examples based on the RIPER-5 model, a specification-driven design approach, the talk demonstrates how to merge the potential of AI with already established practices to create balanced, reliable and developer-centered workflows. Participants will walk away with a practical roadmap to cut through the hype, maintain human oversight and integrate AI in ways that deliver real value.

Many AI initiatives fail because the data chain is fragmented, slow and lacks machine-readable governance, making it difficult to create high-quality data products. This session shows how the Thoughtworks team accelerates this process from weeks to hours with AI Building for AI, already used by a life sciences client. The platform, Zap, uses specialized LLM agents to execute the full data product lifecycle, transforming teams into strategic reviewers focused on value and governance. The content presents the agent-oriented architecture and the evaluation frameworks that enable speed, trust and scalability for enterprise AI.

For years, LATAM Airlines operated with an agility-based model focused on speed and delivery, but as the market evolved, the organization hit a ceiling. Teams shipped features efficiently but without generating long-term strategic value. This session presents how the Thoughtworks team supported LATAM Airlines in transitioning from an agile structure to a true product operating model, and how product operations became central to that shift. The content covers practical steps and frameworks used to reshape teams, roles and governance toward outcome-oriented work, the leadership approach needed to guide people through resistance and change, and the concrete results that prove the impact of a mature product practice.

At a major Canadian fintech, an AI-First Software Engineering Transformation program is redefining how teams build and deliver software. The initiative introduces a new framework and a set of practices that integrate AI across the entire software development lifecycle while addressing the cultural and organizational challenges that come with large-scale change. The talk explores the program’s motivation, phased implementation and the measurable impact on engineering productivity and effectiveness.

This panel brings together different perspectives on the inclusion of people with disabilities, featuring Carolina Ignarra, founder of Talento Incluir, leaders from the Globo account at Thoughtworks and professionals from Globo itself. The panel explores how inclusion initiatives move beyond intention and turn into real practice, showing how accessibility and representation fuel innovation inside organizations. The conversation highlights real examples of collaboration between consulting teams, internal leadership and clients to build work environments where technology truly serves everyone.

In a world increasingly mediated by algorithms that influence everything from content consumption to decisions about credit, work and opportunity, the central reflection is understanding when these technologies start reproducing social bias and inequality. This presentation shares experiences on ethical, responsible and inclusive use of AI, focusing on real impacts, automation dilemmas and global guidelines for transparency and fairness. Through practical examples, it highlights initiatives that aim to balance innovation and ethics and invites the audience to reflect on whether technology is serving people or people are serving technology.

Caixa Econômica Federal sought to modernize its credit risk model pipelines and, when assessing its needs, identified an opportunity to go beyond Data work and build an infrastructure architecture based on Data Mesh. After alignment with the Cloud and Security teams, the project moved forward with an MVP capable of significantly accelerating model implementation, where the value of starting directly in a Data Mesh setup was quickly recognized. Collaboration between multiple teams enabled the creation of a solid foundation for future evolution, generating relevant learnings to share.

How do you drive innovation in a regulated, high-impact sector that often resists change? This session shows how the Thoughtworks team worked with LATAM Airlines’ Maintenance area to build an Experimentation Lab and apply a framework that brings together Data, Design and AI to increase leadership confidence, accelerate time to market and boost the value of digital transformation by more than fifty percent.

AI is accelerating organizational change, yet many companies still struggle to turn potential into real productivity gains. Agentic AI enables true collaboration between people and intelligent systems, moving beyond isolated task automation. In this talk, we’ll outline the main barriers to delivering value with AI and introduce a business architecture that closes the gap between ideas and impact. Using real examples, we’ll show the platform foundations needed to scale AI responsibly and prepare the organization for a truly AI-ready future.

The vision of an autonomous cloud capable of provisioning, scaling, protecting and optimizing itself is beginning to take shape with AI-driven operations. This session shows how artificial intelligence is transforming the foundations of GitOps, observability and automation to create adaptive, self-managing platforms, and how AIOps, machine learning and reinforcement learning are used to detect anomalies, predict failures and optimize workloads in real time. The content clarifies what is viable today versus what is still experimental and presents a practical path connecting platform engineering with new AI capabilities to prepare teams and platforms for intelligent infrastructure.

Hours spent writing documentation that nobody wants to maintain? This talk shows how AI can turn slow, manual processes into smart, fast and consistent workflows. Through the MDX-GDocs case study, it demonstrates how an agent built with Gemini and the Model Context Protocol (MCP) can convert text or Word files into MDX documentation with Docusaurus components, ready for automatic integration into the repository through Pull Request. Beyond the technical demo, it shares insights on applying AI ethically and practically within engineering teams, reducing friction, improving developer experience (DX) and freeing time for what really matters: building.

This panel brings together industry leaders to discuss the transformative potential of AI in both technology and society, connecting strategic perspectives on AI Strategy, AI First Software Delivery, Agentic AI and how AI is already impacting operational efficiency and top-line value creation. The conversation explores opportunities and risks, the role of organizations in responsible AI adoption and practical paths to scale impact in complex environments. With experts from different sectors, the panel highlights how AI is reshaping business models, engineering capabilities and organizational decision-making, offering valuable insights for anyone navigating this next technological era.