Webmentat
Webmentat

The technology behind Webmentat

Founders choosing a foundation for a real SaaS product deserve to know exactly what it is built on, not a vague claim about "modern technology."

This is the stack, the infrastructure, and the engineering approach behind Webmentat products, the Framework, and this website.

Core application stack

Boring, well-known tools chosen on purpose, not the newest thing available.

React Router Framework Mode

Explicit routes, server rendering, data loading through loaders, and mutations through actions, in one full-stack application model. Not Remix.

React

The component and rendering model behind every Webmentat interface.

Node.js

The standard server runtime and development environment.

Vite

The development server and production build system.

Plain JavaScript and JSX

No TypeScript. A deliberate choice to keep the codebase direct and readable, without a second language layer between the code and the person reading it.

Tailwind CSS and DaisyUI

Utility-based styling with semantic theme tokens, so every interface shares one design system instead of a new one per application.

What this website runs on

The concrete providers behind webmentat.com today, not a hedge about what might be used somewhere.

Render

Application hosting and runtime infrastructure for the Webmentat website.

MongoDB Atlas

Stores submitted inquiry and contact form data.

AWS Secrets Manager

Holds application secrets outside the codebase, injected per environment.

Fastmail

Sends transactional email through Fastmail's JMAP API, including form notifications.

What the Framework supports

Webmentat products are built around provider-independent capabilities, with explicit provider implementations behind them. A product uses a provider because it needs the capability, not by default.

Stripe
Payments, billing, subscriptions, and connected-account payment flows.
Cloudinary
Application media storage, transformation, and delivery.
Google and Apple
Sign-in, plus Google Maps and location functionality where a product needs it.
OpenAI
AI-assisted product features such as content assistance, structured generation, and translation.

Development and delivery

Docker
Reproducible development and deployment environments, so the same application artifact moves through every environment with fewer differences.
GitHub
Source-code hosting, development collaboration, automated workflows, and package and container-image distribution.
Caddy
Local development routing, HTTPS termination, and reverse proxying.

Human decisions, AI-assisted engineering

Webmentat uses AI throughout architectural exploration, implementation, debugging, documentation, and review.

Final architectural decisions remain human. No AI-generated code becomes part of the accepted codebase without careful review. Code is not accepted merely because it resolves the immediate task. It must respect the architecture, reuse existing systems, avoid duplicated logic, remain understandable, and solve the problem without unnecessary complexity.

AI accelerates the work. Human judgment protects the architecture.