# Building ambitious software — Jonathan Kelley, Dioxus Labs & Cognition

## Executive summary

Jonathan Kelley discusses building ambitious, cross-platform software using Rust (Dioxus), detailing the challenges of maintaining high quality across a complex stack. He highlights how AI coding agents have dramatically accelerated development—allowing the team to implement features like deeply integrated Kotlin and Swift plugins in weeks—but notes that this shift changes the focus from writing code to maintaining architectural quality. The core message is that while code is now cheap, architectural integrity and quality remain the primary concerns for ambitious projects.

## Key takeaways

- The Shift in Software Engineering Focus: With AI agents handling the low-level development burden (e.g., fighting the borrow checker), the job of a software engineer shifts from writing code to architecting elegant solutions and thinking about system evolution.
- Code Quality vs. Code Quantity: The speaker asserts that 'code is cheap now, but quality is not.' Maintaining a high quality bar across the entire codebase is critical for developer experience and business stability.
- AI Agents as Knowledge Accelerators: Coding agents excel at 'knowledge problems,' such as sifting through thousands of pages of documentation, reverse engineering APIs, and recalling deep technical specifications (e.g., CSS specs), accelerating tasks that previously took years of hand-written effort.

## Technical details

- Dioxus Framework: A cross-platform app framework written in Rust, aiming to allow development using HTML and CSS as the markup language, supporting web, iOS, and Android apps.
- Blitz Rendering Engine: A lightweight, custom web browser-grade CSS and HTML rendering engine developed for Dioxus. It is noted for its small bundle size (less than 5 MB) and low runtime memory consumption (less than 50 MB), compared to Electron apps.
- Subsecond Hot Reload: A generic hot reload engine for Rust, C, and C++ that watches code for edits, recompiles changed parts, and patches the running application in place in approximately 100 milliseconds.
- Rust Development Acceleration: AI agents are capable of handling the 'borrow checker' and 'edge cases' in Rust, effectively reducing the language's learning curve from a development burden into a feature.
- Build System Integration: The team successfully implemented deeply integrated Kotlin and Swift plugins for Dioxus, a feature that previously required years of manual development effort, achieving it in weeks using coding agents.

## Practical implications

- Build engineers should prioritize architectural design and maintainability over raw feature velocity, as the foundation dictates the ceiling of future development.
- Leverage AI agents for automating tedious, high-volume tasks like updating documentation, verifying release checklists, and backporting bug fixes to maintain a rapid, consistent release cadence.
- Focus testing efforts on building robust fuzzing harnesses to test foundational software against adversarial and malformed inputs, as agents struggle with writing the 'right' test cases.
- The complexity of modern build systems (e.g., cross-language plugins) can be significantly reduced by treating the language learning curve itself as a feature that AI can manage.

## Topics

Rust, Cross-Platform Development, Build Systems, Hot Reloading, AI Coding Agents, Software Architecture, CI/CD, Jonathan Kelley, Dioxus Labs, Dioxus GitHub

Source: https://www.youtube.com/watch?v=H7vFrcNWXzs