Nawal discusses using LLMs for migrating component libraries at Zalando, highlighting challenges and complexities. Discussing a recent migration problem at Zalando involving unification of component libraries and challenges in automating migration process. Encountering limitations with codemods, leading to exploration of LLMs for automated migration processes. Conducting iterative experiments to enhance LLM accuracy in component migration. Building a CLI tool for component migration. Using AIDAR for automating git commits & maintaining LLM accuracy. Recommendation for team closeness to code, successful complex transformations by LLM, challenges with hallucination and ignoring instructions. Lessons learned include field addition, non-deterministic outputs, slow responses, project constraints, and tool evaluation. Evaluating tool enhancements, PROMPT best practices, core approach, improved tools, LLMs for simple migrations, and tool limitations with multiple file changes.

Jamin discusses the shift from using form libraries to native web APIs in React development, emphasizing the benefits of exploring web-native APIs before relying on libraries. The talk covers topics such as Constraint Validation API, React 19 form actions, and CSS pseudo-classes for form building without libraries. It highlights real-time validation, simplified form submission, and custom error handling in React forms. The importance of customizing error messages, setting up custom error handling, and optimizing form performance are also addressed, along with challenges and recommendations for utilizing web native APIs in form development.

Steven discusses the role of taste in software development and its implications for the future of software engineering. The focus is on Cognition's product Devon, an AI software engineer, augmenting the team's capabilities and achieving significant outcomes. Devon's extended responsibilities include testing, reviewing, and monitoring features, with a prediction of English becoming the primary source of truth in software engineering. The transition to English as the primary language for software development is envisioned, emphasizing stability, history-awareness, and spec-driven development.

Eric Handshed introduces spec-driven development, emphasizing its benefits in creating high-quality code by writing specs before coding. Spec-driven development involves structured specifications written before coding, enhancing collaboration with language models and coding assistants. The approach guides developers and ensures code quality. Hanchett stresses the importance of guiding large language models with context to ensure they follow the right direction amidst evolving software requirements and paradigms. Be cautious with context in spec-driven development to avoid overwhelming or underwhelming large language models. Use skills to guide coding agents effectively and maintain a human-in-the-loop approach for code review in the spec-driven development process. AWS history lesson leading to Kiro release with CLI option and emphasis on vibe and spec mode for larger projects and complex features. Release of spec-driven development process, manual method without Kiro, Speckit usage in Qro for new and existing projects. Requirements document creation, design document tips, implementation phase tasks with property-based tests, and model context protocol for data source connections. MCP maturation, integration with project management, demo of QIRO IDE for project creation, MVP creation, and property-based tests for requirements satisfaction. Creation of a movie website starting with a design document, generation of property tests using FastCheck in TypeScript, development of requirements document, task list update to create an MVP, implementation of MVP features, and examination of property-based tests.

Rudhankariyam discusses AI-driven coding to develop self-maintaining AI tools. Introduced an open-source project, App Store Connect CLI, as a case study for AI-driven coding. Implemented automation for App Store publishing using agents and APIs. Managed fewer PRs in bulk, used codex and cursor cloud automations efficiently. Added aliases for hallucinated cloud code commands, improved descriptions and examples. Implemented separate auditor agent for rigorous code review. Emphasized human decision-making for product direction and initial rules in automation strategies. Evaluate model quality based on error rates and user expectations. Highlighted the importance of human judgment in software development amidst AI advancements.

David, creator of React Bits, shares insights on designing standout UI components, emphasizing the importance of distinct interfaces without being noisy. ReactVids focuses on creating reusable visual moments for developers, ensuring standout UI accessibility. Community-driven evolution of React Bits prioritizes customization, usability, and key UI qualities. Magic Rings exemplifies the balance and purpose needed in UI components. The versatility of Magic Rings showcases the importance of a flexible visual system. Exploring shaders enhances the control and visual richness of Magic Rings. Leveraging shader uniforms enables dynamic visual effects in React components. Real creative controls in component design emphasize meaningful customization. Strategic props play a crucial role in reusable creative component development. The philosophy of React Bits focuses on easy adoption, customization, and providing a strong starting point for distinct interfaces. ReactVids aims to create memorable, intentional, and unique user moments in UI design.

The case study from Zotoo highlights a new fourth platform reshaping the way applications are developed. Meeting user needs across devices presents challenges and opportunities in app development. Scaling challenges and technological evolution involve adopting new technologies like PWA and TWA. Innovative technology approaches with React Native and transitioning to a universal system framework are key focus areas. The organization is evolving its structure towards a stream value organization, optimizing domain focus, and expanding reach with future plans for smart TVs and mobile devices.

The Talk covers various aspects of deception and security in AI and data systems. It discusses the historical use of misinformation during World War II to influence decision-making. AI model poisoning through malicious data input and the challenges of ensuring data integrity in real-time pipelines are highlighted. The importance of peripheral information attacks, data contracts, stream governance, and anomaly detection in maintaining AI system security is emphasized. Strategies for protecting data integrity, detecting anomalies, and choosing appropriate AI models are also discussed.

The Talk delves into the migration journey of over 500 micro frontends, emphasizing the use of AI agents for efficient migration. It discusses the challenges of handling manual PRs and code mods at scale, highlighting the importance of reliable tooling. The balance between automation and human judgment is crucial, with deterministic lanes handling basic tasks and human intervention reserved for complex issues. Scalable agent orchestration, workflow automation, safe PR launches, and stable testing are key themes, emphasizing the collaboration between machines and human judgment for successful migrations.