8080.AI

8080.AI is a product of F(x) Data Labs, Inc that uses a team of specialized AI agents to turn a plain-language idea into a running application. Users describe what they want, approve requirements, designs and plans at review gates, and agents carry the work through code, tests, security scans and deployment. It is available on the web, with a VS Code extension and iOS and Android apps linked from the site.

What It Is

8080.AI is an agentic software-building platform. A supervisor agent decides which specialists to involve, such as a system requirements agent, user flow planner, design agent, system architect, project manager and frontend, backend, DevOps, QA and AI engineer agents. The outputs include a requirements document, user flows, clickable designs, architecture diagrams, a task plan and a deployed app, with a codebase that merges into the user's repo by pull request.

How the Nine-Step Process Works

The site describes nine steps: share an idea, approve requirements, approve user flows, approve designs, edit visually, approve the blueprint and task plan, let agents build in parallel, scan each task for security issues, and open the live app. Each task runs in its own isolated cloud workspace through a six-phase pipeline of setup, environment, branch, code, tests and staging. Before staging, tasks are scanned for leaked secrets, vulnerable dependencies and unprotected API routes. Finished tasks deploy to a project-specific Kubernetes namespace with a managed backend including API, database, auth and storage.

Humans and AI Together

Projects can be shared across organizations with owner, admin and member roles. Several developers can work from one Kanban-style board alongside AI engineers, with list, sprint, Gantt and phase views. Work can be assigned to humans for decisions or reviews, after which agents pick the thread back up. Developers can use a full VS Code editor in the browser or the desktop extension.

Model Approach

The company says it uses a routed attention approach (dynamic dense attention, O(n)) that retrofits existing open models to keep cost flat over long context, and publishes a proof page. These performance figures are vendor claims.