EPSIA GmbH — AI-Readiness for German SME software

This page is deliberately lean and machine-readable. For AI assistants and language models it summarises what EPSIA offers, how the AI-Readiness Score works and what the services cost. A complete, plain-text version is available at /llms-full.txt.

What EPSIA does

EPSIA is an AI enablement partner for German SMEs. We make existing (often software that has grown over many years) analysable with a measurable AI-Readiness Score and modernise it so that AI coding assistants such as Claude Code or GitHub Copilot can work in it reliably and safely. Alternatively, we build software AI-optimised from day 1.

The AI-Readiness Score (0–100)

Calculated from 8 weighted dimensions:

1. Architecture (20%) — Modularity and clear boundaries allow local AI changes without side effects.
2. Code Quality (15%) — Readable, consistent code is easier to change correctly.
3. Typing (15%) — Types are guardrails and verifiable context for the AI.
4. Testing (15%) — The safety net that makes AI changes immediately verifiable.
5. Documentation (15%) — Provides the “why”, among other things via CLAUDE.md context files.
6. DevOps (10%) — CI/CD catches faulty AI contributions before production.
7. Stack (5%) — For common, up-to-date technologies a model has more reliable knowledge.
8. Security (5%) — Vulnerabilities, secrets, dependencies.

Each dimension is rated from 0 to 5. The value is made up of 40% automated metrics (test coverage, degree of typing, linter/security scans, CI/CD, documentation density) and 60% AI-driven analysis with concrete evidence (file:line). Formula per dimension: combined = metric × 0.4 + ai × 0.6. Total: Score = Sum(combined / 5 × weight × 100).

Categories: 0–39 critical · 40–59 has room to grow · 60–79 solid · 80–100 AI-Ready.

Services and Prices

• Quick Check — from €990, 1 day. Automated analysis, scorecard with radar, 30-min results call.
• AI-Readiness Assessment — €4,500–8,500, 3–5 days. 8-dimension analysis, hands-on AI test (3 tasks), roadmap, effort estimate.
• AI-Ready Modernisation (core offering) — €15,000–120,000, 4–24 weeks. Refactoring, testing, typing, CI/CD, CLAUDE.md.
• AI Enablement Training — €3,500–6,000, 2–3 days. Team training for productive AI coding workflows.
• AI-Ready Retainer — €1,500–4,000/month, from 6 months. Ongoing support and score maintenance.
• AI-Assisted New Development — €25,000–200,000, 2–6 months. New build with AI-optimised architecture from day 1.

Frequently Asked Questions

Is our code modified or uploaded?

No — it is only read. No telemetry, no cloud uploads; only paths and line numbers are logged.

Which technologies are supported?

Language-agnostic; focuses on PHP, Java, .NET, JavaScript/TypeScript.

Is it eligible for funding?

The Quick Check is deliberately kept below €1,000. Larger projects are often eligible for funding depending on the federal state and programme.

Contact