PORTFOLIO

AI Engineer · Founder · Edge Systems Specialist · Based in Germany

Available for Collaboration

Supercharging
Human + AI

Driven by radical optimism. I build systems that leverage artificial intelligence for good, helping shape a better world through robust, production-ready technology.

Multi-Agent Systems
Edge AI & ML Ops
Production Backends
Data Orchestration
VIEW ARCHITECTURE

Engineering
Execution

Building systems that survive real load, real users, and the full entropy of production environments.

01

Robust Backend Architecture

02

Edge AI & HW Optimization

0+

Platform Users

0+

Years Engineering

0+

Systems Shipped

0+

Countries Reached

End-to-End Capabilities

INTELLIGENCE
PORTFOLIO

Featured Projects02

3d.storsko.com

Founder & AI Engineer

Scaled an AI platform to 3,000+ active users. Architected multi-agent orchestration via LangChain/LangGraph with a production backend (FastAPI, RabbitMQ, PostgreSQL).

Stack
LangGraph
Users
3,000+

Hensoldt

Intelligence Systems Engineer

Developed advanced edge AI infrastructure for defense applications. Presented benchmarks and model evaluations to technical leads. Optimized on NVIDIA TensorRT AGX Orin.

Hardware
NVIDIA AGX
Tech
C++ / ML SOC
Career Track

EXPERIENCE
TIMELINE

    Storsko

    Founder & ML Engineer

    Scaled an ML platform to 3,000+ users. Built 3D model generation pipelines cutting inference costs by 70%. Architected edge infrastructure via LangGraph, RabbitMQ, Docker, and Supabase.

    • 3D Generation
    • Pipelines
    • Distributed Systems

    Hensoldt Optronics

    Dual Study Student AI

    Evaluated ML pipelines for edge hardware in autonomous defense systems. Developed a microcontroller-based ground surface detection for navigation systems and automated ID generators.

    • Edge AI
    • Autonomous Tech
    • TypeScript/JSON

    DHBW University

    B.Sc. Computer Science (AI)

    Built a high-performance neural network training framework from scratch in C with Python bindings. Thesis on ML model robustness under multispectral sensor misalignment.

    • C / PyTorch
    • Deep Learning Theory
    • Research
Technical Depth

SKILL
MATRIX

Full-Stack AI Engineering
AI & Agents
  • LangGraph / LangChain
  • Multi-Agent Orchestration
  • RAG Pipelines
  • LLM Fine-Tuning
Backend
  • FastAPI / Python
  • PostgreSQL / RabbitMQ
  • REST & GraphQL
  • Microservices
Edge AI
  • NVIDIA TensorRT
  • AGX Orin Platform
  • C++ Inference
  • Model Optimization
Frontend
  • Next.js / TypeScript
  • React Ecosystem
  • TailwindCSS
  • Three.js
DevOps
  • Kubernetes / Docker
  • CI/CD Pipelines
  • Cloud Infra (AWS)
  • Monitoring
Security & Ethics
  • AI Safety Research
  • Responsible Deployment
  • Privacy by Design
  • Compliance
Philosophy

EVERY SYSTEM IS
A PRODUCT OF
DELIBERATE
CHOICES.

I don't build demos. I architect systems that survive contact with reality — with users, load, edge cases, and organization entropy. Production is where design is truly tested.

M
Milan Kiele
Intelligence Labs

Building the systems that power tomorrow's AI — one deliberate commit at a time.

GitHub

@MilanKiele

LinkedIn

Milan Kiele

X / Twitter

@milankiele

Email

mail@milankiele.com