Modern Stack, Edge Infrastructure & Effective AI
Building durable, ultra-low latency software systems with minimal maintenance, simple frameworks, and highly productive dev teams.
Cloudflare Edge Infrastructure
We focus on core operational benefits: fast startup times, low maintenance overhead, global distribution, fair pricing, and strong ecosystem support. We favor Cloudflare Workers because it delivers serverless computing directly at the edge with native WebAssembly (WASM) support—allowing applications to run at peak speed worldwide without managing servers or containers.
Web Components & Native Standards
We believe two major benefits of React have been diminished: HTML components and the Flux one-way data path architecture. By seeking to ease the initial adoption curve, React hooks have diminished their component-based approach. Likewise, too often Flux architecture isn't implemented because it requires too much effort up front.
In contrast, it's better to start with a lightweight framework and leverage Standard Web Components for HTML components. It's simple to use, simple to understand, less complicated to build, easier to debug, and has fewer moving parts to maintain.
Lightweight & Pragmatic Frameworks
We fundamentally believe web frameworks have become too complicated and we need to adopt lighter, simpler frameworks with minimal packaging and build requirements. More complex frameworks (e.g. React, Angular, Vue, Svelte) serve a role, but we argue that their primary benefit is not about technical elegance, but more about team recruitment and skill cohesion.
Basically, any framework that is lightweight, elegant, and supports building standards-based websites and applications, we will use. Frameworks we like include: htmx, Alpine, Astro, Fixi, and Unpoly—and for mobile: Capacitor and NativeScript.
Rust & AI-Accelerated Engineering
Leveraging Antigravity/Gemini, Claude, and modern AI pair-programming agents to focus on domain modeling, systems boundaries, and architectural precision allows small, highly focused dev teams to deliver maximum output with minimal overhead.
We favor Rust because it is explicit and catches errors at compile time. We believe Rust is the ideal language for AI-assisted engineering because it leverages mutual strengths while protecting against the other's weaknesses.
AI generates code rapidly, flattening Rust's learning curve so an engineer doesn't need complete mastery of ownership models and lifetimes to be immediately productive. Enduring slow compile times is far less of a problem when AI writes the code so quickly.
Conversely, when AI hallucinates or gets the architecture wrong, Rust's explicit nature, strict compilation requirements, and actionable error messages provide a crucial safety net. Rust catches flaws early and forces clarity of design and architecture. More often than alternatives, if Rust compiles, it generally does what you want it to do. Combined with best-in-class runtime performance, minimal resource consumption, and the ability to compile to WASM, we believe Rust is very well positioned within the AI world.
Get in TouchDiscuss a Project or Technical Collaboration
Have a complex software challenge, edge architecture inquiry, or small business project? Send a message directly.