metrix – Strategy Analysis Across 17 Business Frameworks

THE IDEA

metrix – Strategy Analysis Across 17 Business Frameworks

Type: Software Status: Beta Tech Stack: React 19, TypeScript, Vite, Tailwind CSS v4, React Router, Node/Express, better-sqlite3, tsx, jose, bcryptjs, marked, Puppeteer, Docker, nginx, multi-LLM (Anthropic/OpenAI/Google/xAI/Kimi/DeepSeek/MiniMax/Ollama/vLLM/LM Studio) Problem Statement Looking at a business strategically means working through a whole series of frameworks — market sizing, problem prioritisation, offer design, canvases, a financial model, risks. Done by hand that is slow and rarely consistent, and buying it in as consulting is expensive. What is missing is a tool that walks the frameworks in one pass, keeps the result exportable, and can run entirely on local models when the input is sensitive. Description metrix turns a handful of business inputs (niche, industry, target group, budget, bottleneck) into an end-to-end strategy analysis across 17 established business frameworks — from market sizing (TAM/SAM/SOM), problem prioritisation and offer design through Business Model, Lean and Value Proposition Canvas and the 55 St. Gallen patterns to a live-calculated financial model, an OKR action plan, a risk register and a cross-framework Business Health Score. The app is provider-agnostic, treating cloud APIs and local models as equals (including auth for local model servers), generates primarily in one language (German or English) and translates the second on demand, and can feed real data into any framework as evidence. An optional Node/Express backend with SQLite adds a persistent history (with version chains and workspaces), a REST/RPC API, outbound webhooks and two auth layers (service token, optional Keycloak or local browser login), so external orchestrators can trigger analyses programmatically; without the backend everything runs standalone in the browser. Notable extras: a Council mode that runs several models in parallel plus a moderator synthesis, an experimental Synthetic Market with up to 50 AI personas as a test audience (including A/B comparison, a word-of-mouth round and 1:1 chat), and export as PDF, Markdown, JSON or pitch deck. A broad, security-focused test suite (SSRF hardening, auth hardening, PDF render hardening) and a restrictive CSP in the deployment underline how close to production it sits. Use Case Anyone who wants to examine a business or an idea strategically enters a few key figures and gets a complete, exportable strategy folder from the AI — financial model, risks and investor pitch included — entirely locally and without consulting fees if desired. Link: https://github.com/rawk7000/metrix (private repo) The running analysis: 17 frameworks, market sizing and demand trends in one view. New analysis: enter the key figures — or have the fields filled automatically from a brain dump. History: reopen, export or regenerate saved analyses. Synthetic Market: testing an offer against a population of AI personas — directional, not predictive. Quick Tools: standalone generators for repurposing, content calendars and hooks. Settings: cloud providers and local models are offered on an equal footing.