Documentation

Architecture, quick start, owner playbook with visuals, API reference, CLI truth table, and links to the Markdown handbooks in docs/.

AI-Factory v2.1

AI Company Platform · Single Docker Stack

Production-grade stack in one container (FastAPI + Next.js). One multi-agent pipeline for everything: autonomous mode feeds market research and generated ideas; on-demand mode starts from the customer's phrase — same stages, same QA gates. Typical deliverables are share-ready landings; crypto checkout and sandbox preview ship with the bundle.

Core Capabilities

Twelve specialized pipeline roles in the admin roster: Analyst, PM, Methodologist, Architect, Designer (UX), Developer, QA, Security, DevOps, Marketing, Sales, and Evolution Analyst — plus Director as a meta-agent; `ui_experience` in architecture is the binding UX brief for the Developer
Multi-LLM routing with failover — local Ollama, DeepSeek, Together, Groq
Director AI — scheduled analysis, decisions queue, reports
Crypto storefront — USDT/USDC (Base, Arbitrum, Ethereum, Solana); default list ~$4.99 USDT when sales artifacts omit price
Glass storefront & admin — animations, responsive layout
Enterprise-minded ops — secrets paths, audit hooks, sandbox isolation
CLI companion for operators (`ai-company` where installed)
Evolution loop — telemetry-driven improvements

At a glance

Compose

Deploy model

Demo + QA

Quality gates

~$5

Pricing default

Full

Stack depth

Single Container

Everything runs in one Docker container (Ubuntu 24.04, Python 3.12, Node.js 20). No external dependencies required.

Canonical handbook (Markdown)

For printable diagrams (Mermaid), full scenario tables, and pitfalls see docs/owner-guide.md in the repository — kept in sync with this tab.