Freelance Consulting
I mentor R&D teams through the full machine learning lifecycle—from tailored models to sustainable production systems.
The Challenge
- Promising models stall as prototypes and never make it into production.
- Codebases accumulate technical debt, slowing development and blocking progress.
- Project goals and architectural design are unclear, leading to wasted effort and rework.
- MLOps practices are missing, so even working models can't be sustained over time.
A Better Way
I help your team avoid these pitfalls by combining deep machine learning expertise with software engineering and MLOps best practices.
My approach delivers concrete results while ensuring long-term sustainability:
- Dive into your data and domain to understand the problem in depth.
- Experiment with tailored ML models and demonstrate feasibility through proofs of concept.
- Pair with your team to refine prototypes into clean, maintainable code.
- Build robust data pipelines, and set up monitoring and retraining workflows.
- Optimize architecture and eliminate technical debt so the team stays unblocked.
- Mentor developers so they gain the skills to own and maintain solutions after I'm gone.
The outcome: projects move beyond prototypes to provide lasting impact while your team gains valuable skills.
What I Offer
I work with teams in a flexible, hands-on way—providing immediate value while setting them up for long-term success.
Core Offering: Long-Term Mentorship
The main way I work is by joining your team part-time (around 20h/week for 6–12 months). I support projects end-to-end: from early experimentation through production deployment and beyond. Along the way, I level up your team through code reviews, architectural guidance, and hands-on help where needed.
Entry Packages
- Use Case & Concept Workshop (1-2 weeks)
For teams beginning a new ML or data-driven project. We define the use case, clarify scope and requirements, and design a solid architecture from the start. The outcome is a shared plan that avoids wasted effort and sets the foundation for clean, production-ready development. - ML Systems Audit & Roadmap (2-4 weeks)
For teams with prototypes that haven't made it into production. I review your codebase, data pipelines, and ML workflows to identify weaknesses in architecture, code quality, and maintainability. You receive a clear roadmap for refactoring and strengthening the system so it can scale reliably into production.
Why Me
- PhD in machine learning with strong expertise in Python, data science, and scientific computing.
- Years of experience in the process industry and manufacturing, giving me insight into the challenges of applying ML in real-world industrial and R&D contexts.
- Track record of helping teams take ML projects from messy prototypes to production systems with impact.
- Deep knowledge of both ML algorithms and modern software engineering, MLOps, and architecture best practices.
- Focus on mentoring: your team remains the owner of the solution and grows self-sufficient.
Ready to Work Together?
Whether your team is starting a new ML project or struggling to move a prototype into production, I can help you deliver results that last. Send me an email at hey@franziskahorn.de to discuss which entry package makes sense as a starting point.