Zum Hauptinhalt springen

Guten Tag

Dr. Florian Drechsler

Mathematician, Software Engineer, and AI Developer Based in Leipzig

Dr. Florian Drechsler combines mathematical rigor with modern software engineering. Florian Drechsler develops intelligent systems for predictive maintenance and energy optimization at TreeTek, while building an AI Development Framework that transforms software development through AI-assisted workflows.

View CV

Who Is Dr. Florian Drechsler?

Dr. Florian Drechsler holds a PhD in Mathematics (Dr. rer. nat.) from the Max Planck Institute for Mathematics in the Sciences, a Diploma in Mathematics, and a BSc in Computer Science from Universität Leipzig. Florian Drechsler combines deep theoretical knowledge with practical engineering expertise across machine learning, predictive systems, and full-stack development.

Academic Background

  • • PhD in Mathematics (Dr. rer. nat.)
  • • Diploma in Mathematics
  • • BSc in Computer Science

Professional Experience

Florian Drechsler has over 5 years of experience building production systems, including DPI software at Ipoque, logistics intelligence at NTA Mexico, and ML-based predictive maintenance at TreeTek.

Current Focus

Florian Drechsler builds intelligent maintenance prediction systems and energy network optimizers at TreeTek, while developing an AI Development Framework for automated software workflows.

Florian Drechsler's AI Development Framework for Automated Software Engineering

Dr. Florian Drechsler created the AI Development Framework, a comprehensive system that orchestrates multiple AI agents for automated software development. The framework features ticket-based workflows, specialized agents for implementation, testing, and code review, automated documentation generation, and human oversight checkpoints. Florian Drechsler has successfully deployed this framework in production, completing 54 tickets across 25 AI sessions.

How the AI Development Framework Automates Software Development

Florian Drechsler's AI Development Framework uses a ticket-based workflow where tickets progress through defined statuses: Draft, Ready, Working, Review, and Done. During the Working phase, specialized AI agents collaborate autonomously. The Implementation Agent writes production code, the Test Agent creates and executes tests, and the Review Agent performs quality checks. All agents iterate until quality standards are met, with human approval required only at the final Review checkpoint before merging.

Key Benefits of Florian Drechsler's AI Development Framework

  • Automated Development: AI agents handle implementation, testing, and code review autonomously, reducing manual effort
  • Quality Assurance: Built-in iteration loops ensure code meets quality standards before human review
  • Automated Documentation: The framework automatically generates ADRs, session logs, and deployment guides for every ticket
  • Full Traceability: GitHub issues, labels, and comments provide a complete audit trail for every change
  • Human Oversight: Critical decisions and final approval remain with human developers, ensuring accountability

Projects Built with Florian Drechsler's AI Development Framework

These real-world projects were built by Dr. Florian Drechsler using the AI Development Framework, demonstrating the framework's capabilities in production environments.

fdrechsler.de
Professional personal website built entirely using AI-assisted development through the AI Development Framework. This project served as the first real-world validation of the framework, demonstrating how coordinated AI agents can deliver production-ready systems efficiently. Built with Next.js 15.x, containerized with Docker, and deployed with automated CI/CD. Features include comprehensive analytics (Umami, Google Analytics 4), dark mode support, SEO optimizations with JSON-LD structured data, and full German legal compliance. The project progressed from initial setup to a fully deployed production website through a structured workflow of research, implementation, testing, and review phases, with human oversight at critical checkpoints. Key achievements include comprehensive documentation (ADRs, deployment guides), infrastructure as code, and successful framework bug discovery and resolution.
Active
Next.jsTypeScriptDockerCI/CDProduction
CardCroc
Mobile flashcard learning app with Spaced Repetition for iOS and Android, built with Flutter. CardCroc helps users learn efficiently through the scientifically-proven SM-2 algorithm. The app features Firebase Authentication (Google & Apple Sign-In), full internationalization (DE/EN/ES), a vocabulary data pipeline producing 20 learning decks with ~10,000 cards, and a freemium model with Pro themes. Developed entirely using the AI Development Framework across 43 releases (v0.1.0 to v0.16.0), demonstrating the framework's capability to deliver production-ready cross-platform mobile applications.
Active
FlutterDartiOSAndroidMobile
TradingOctopus
AI-driven trading platform for collecting, analyzing, and acting on financial market data at scale. TradingOctopus implements a modular microservices architecture with a high-performance FastAPI REST API, TimescaleDB for time-series storage, and a multi-source data mining service with real-time market data collection. The platform features comprehensive domain models (instruments, listings, venues, OHLCV candles, fundamentals, corporate actions), event-driven architecture with flexible JSONB-based financial events, data provenance tracking, and a Yahoo Finance collector with multi-timeframe support. Built entirely using the AI Development Framework with 3 rounds of test quality audits ensuring production-grade reliability, demonstrating the framework's capability to deliver complex financial data infrastructure.
Active
PythonFastAPITimescaleDBDockerTrading
ResearchRaven
AI-powered research-to-content pipeline that processes knowledge in three stages: Research (linear) → Knowledge Organization (hierarchical) → Content Generation (linear). ResearchRaven automates the entire journey from source collection to publication-ready content like blog posts, newsletters, and LinkedIn posts. The architecture is grounded in communication theory (Shannon, Saussure, Miller) and the principle that hierarchical knowledge must be linearized for human consumption. Features parallel multi-agent research, a hierarchical knowledge base with MCP server interface, style patterns for multiple output formats, and a humanize post-processing pass. Built entirely using the AI Development Framework with Claude Code.
Active
PythonBashMCPAI PipelineContent
TicketFalcon
Chat-based AI assistant that helps teams write better tickets. TicketFalcon connects to company knowledge sources (Confluence, Git repositories) via API, lets users describe requirements in natural language, and uses LLMs to formulate well-structured tickets — which are then created directly in the target ticket system. The desktop application supports pluggable LLM providers (Claude, OpenAI, Mistral, Ollama) and pluggable integrations for ticket systems and knowledge sources, with Jira as the first integration target. Built using the AI Development Framework with Claude Code.
Active
Desktop AppAIJiraChatProductivity

Florian Drechsler's Blog on AI Development and Software Engineering

Dr. Florian Drechsler writes about AI-assisted software development, machine learning, mathematical approaches to engineering, and building intelligent systems.

How to Contact Dr. Florian Drechsler

Dr. Florian Drechsler welcomes inquiries about collaboration opportunities, project ideas, and professional connections. Florian Drechsler is open to discussing new opportunities in software engineering, machine learning, and applied mathematics.