25+ Years in Business.
Now with AI.

For over 25 years, I've worked with companies that deliver results, not just promises. What I've built works - four live projects, real automation, measurable outcomes.

Discuss your project
Scrollen
Results
25+ Years.
One Business.
Four Projects.
Real Results.
Measurable

What drives my work

25+

Years running businesses

9 → 1.5h

Weekly admin time after AI automation

Content output tripled with half the effort

I'm a practitioner. I've built AI systems - for my own business, under real cost pressure, with real processes.

Since 2002, I've run mindmelt as an owner-managed agency in the Rhine-Main region. Clients like Schott, Zeiss and Helaba have trusted me for years - on projects where precision mattered and mistakes had consequences.

The four projects below are built in-house, running productively every day, and developed from genuine requirements of a real business. From back office through content production and SEO to completely automated ticketing systems - all digital automation that actually works.

Case Study 01
Back office.
ERP.
Automated.

Back Office Automation & ERP Integration - mindmelt Frankfurt

The Challenge

Until 2024, mindmelt ran typical isolated systems: proposals in Word, invoices in DATEV, customer data across various spreadsheets, communication on three channels. Every task was manual. Administrative overhead ran to eight to ten hours per week.

What We Built

A complete back office as a bespoke web application: invoice generator with automatic numbering, expense receipt creation and centralised customer database. The system generates ZUGFeRD-compliant invoices as PDFs - directly importable into DATEV and other accounting systems. Every process from input to finished document is automated.

Results

Administrative tasks: from ~9h to ~1.5h per week. Invoice creation: from 45 minutes to under 5 minutes. Invoice error rate: zero. The system has been running live since March 2025 - without a single manual correction.

Stack

  • Python / Flask backend with REST API
  • ZUGFeRD/Factur-X for machine-readable invoices
  • LibreOffice Headless for PDF generation
  • Centralised customer DB with automatic matching
Case Study 02
Content.
Pipeline.
Scaled.

Content Pipeline Automation - Social Media & Blog

The Challenge

Content marketing was entirely manual: topic research, briefing, writing, SEO optimisation, formatting for blog and social media, image sourcing. A single blog post with matching visuals took four to six hours of active work.

What We Built

A fully automated content pipeline as a web application. Starting with a topic brief, the system generates: SEO-optimised blog articles (1,200-1,800 words), AI-generated images for each section, automatic layout distribution and direct publishing to REDAXO CMS - with approval workflows at every stage.

Results

Content production cut from 5 hours to under 20 minutes of active time. Output volume tripled. Consistency in tone and SEO parameters dramatically improved. Every article comes with professional AI-generated images in two distinct styles.

Tech Stack

  • Python / Flask for pipeline control
  • Claude AI for text generation and keyword clustering
  • fal.ai FLUX Pro for AI image generation
  • REDAXO CMS integration with automatic module mapping
Case Study 03
SEO.
Backend.
Live daily.

SEO Analysis Backend with AI-Powered Recommendations

The Challenge

SEO reporting across multiple domains required manually combining data from Google Search Console, rank trackers and CMS. Monthly reporting took three to four hours of manual work. Data was always delayed, and action plans were intuitive rather than data-driven.

What We Built

A bespoke SEO dashboard as a web application with ten specialist modules: rankings, keyword research, backlink analysis, broken link detection, on-page audit, SERP features, competitor analysis, local SEO, Google Reviews and Search Console integration. The system automatically detects REDAXO installations and can fix broken links directly in the database.

Results

Monthly SEO reporting cut from 3.5 hours to 15 minutes of review time. Data freshness: daily instead of monthly. AI-powered action plans with impact/effort prioritisation. 24-hour caching minimises API costs whilst maintaining data freshness.

Tech Stack

  • PHP 8.3 with custom routing and PDO
  • DataForSEO API for rankings, keywords and backlinks
  • Google Search Console API (Service Account / JWT)
  • MySQL for historical snapshots and caching
Case Study 04
Tickets.
Images.
Automated.

Ticket System with AI Image Generation & CMS Automation

The Challenge

Requests arrived via email, phone and chat - all in different formats with no consistent structure. Briefs had to be manually compiled, images created separately, and content uploaded to the CMS by hand. Social media posts needed additional approval rounds via email.

What We Built

An email-based ticket system that automatically captures incoming requests, assigns ticket IDs and processes them via AI. Connected to AI image generation (fal.ai FLUX Pro) that creates relevant images on demand. Direct CMS integration with REDAXO: text and images are automatically added to the correct page structure. Social media posts can be published directly from the system once approved.

The Results

Request to finished CMS entry: from two hours down to under ten minutes. Image generation: professional results across six workflow variants (text-to-image, image-to-image, 4K, video). Every step logged with complete conversation history per ticket. Error handling and status updates run automatically via email.

Tech Stack

  • Python daemon with IMAP polling and SMTP responses
  • Claude AI for request processing and content creation
  • fal.ai FLUX Pro for AI image generation across six modes
  • REDAXO CMS integration with media pool connection
  • JSON-based ticket tracking with threading

What can we build for you?

In our initial consultation, we'll show you real numbers from live projects - and what's realistically achievable for your business.
Book your consultation
deutsch english
mindmelt Frankfurt
hallo@mindmelt.de