For SMEs
Replace Excel, paper and no-code processes, modernize grown systems and use AI where it creates measurable relief.
For SMEs & startups
AI engineering, custom software and modernization for SMEs and startups
I build systems that do not stop at the demo but work in everyday operations — from AI features and internal tools to production-ready migrations of vibe-coded prototypes.
Former CTO, now freelance AI engineer: direct collaboration, clear technical decisions and software people actually use.
Segment-Fit
Two starting points, one expectation: reliable delivery.
For startups
Build production-ready MVPs, AI features and CTO-level sparring — including migrating Lovable, Claude Code or other vibe-coded prototypes into reliable apps and websites.
Value Proposition
Why teams work with me
Less hype. More clarity, speed and reliable outcomes.
Production-ready, not demo-only
Operations, maintainability and quality are considered from the first commit.
Direct collaboration
Short paths, honest assessment and no telephone game between strategy and delivery.
Full stack, clear ownership
Frontend, backend, data, deployment and AI considered together.
Business context included
Technology follows the problem, not the hype.
Services
Services for common business situations
Four entry points, one goal: working systems with clear value.
AI systems that work day to day
RAG systems, LLM integrations and multi-agent workflows connected to existing processes, data and interfaces.
Learn more →Software that fits your workflows
Internal tools, web platforms and SaaS products when standard software only covers 80 percent.
Learn more →Productionize vibe-coded prototypes
Turn Lovable, Claude Code or no-code starts into reliable apps and websites — including architecture, data model, deployment and maintainability.
Learn more →Workshops & sparring
Use-case assessment, architecture review or team workshop when clarity is needed before building.
Learn more →Case Studies
Examples that show the collaboration fastest
Concrete systems with understandable impact.
Multi-agent system for a German automotive group
AI-supported analyses with natural language, specialized agents and risk signals.
View project →RAG assistant for internal knowledge management
Production chat assistant with wiki knowledge, legal references and source logic.
View project →Location analysis reduced from days to minutes
Custom analysis app with competitor data, demographics, travel times and AI assessment.
View project →Platform rewrite without standstill
New data architecture, PDF handling and process automation for a grown HealthTech platform.
View project →Process
How collaboration typically works
- 1
Intro call
We clarify goals, status quo, team setup and success criteria.
- 2
Assessment & proposal
You get scope, milestones, risks and a fitting engagement model.
- 3
Iterative implementation
Short cycles, direct alignment and understandable demos.
- 4
Operations & iteration
Clean handover, documentation and optional long-term support.
Engagement
Engagement models and transparency
As much clarity as possible without false precision.
Workshop or discovery
For use-case identification, architecture reviews or AI entry points.
Fixed scope package
For clear goals, scope and acceptance criteria.
Iterative delivery with budget cap
For products, rewrites or projects with learning curves.
Social Proof
What clients highlight
“Niclas understood our idea immediately and translated it into a well-thought-out, stable application. The communication was clear, the implementation fast, and the result fits our requirements exactly. This is what you want software development to be like.”
“With Niclas, we found a partner who quickly and deeply familiarized himself with our technically complex requirements. The technical implementation was clean, the collaboration reliable and uncomplicated. Our platform works exactly as we imagined it.”
“Niclas created a new website for us in no time that achieves our goals and delights our customers. The communication was transparent and professional. We are very satisfied and recommend Niclas to anyone who needs a professional and reliable website.”
FAQ
Common questions before the first call
Is AI right for us yet?
If unclear, we start with assessment instead of a build project.
Do you take over existing systems or vibe-coded prototypes?
Yes, including modernization, no-code replacement, rewrites and turning Lovable or Claude Code prototypes into production-ready apps and websites.
Can you work inside our team?
Yes, independently or embedded as a senior engineer.
Remote or on-site?
Remote across Germany is standard; on-site by arrangement.
How do you charge?
Workshop/discovery, fixed scope or iterative delivery with budget cap depending on the project.
Kontakt
Let’s find out whether it fits technically and commercially
Concrete requirement or early idea: short and specific is enough. I’ll respond with an initial assessment.