Julian von der Goltz

ML systems architect specializing in end-to-end machine learning and deep learning solution development. I build reliable, production-grade systems from concept to deployment with a focus on computer vision and LLM applications. With my formal background in systems engineering and practical experience in AI implementation, I provide a pragmatic, can-do perspective to my teams.

Interested in adding my expertise to your projects? Contact me at [email protected]!

Interests and Specializations

I help companies put AI and automation to practical use. Think searchable knowledge bases with a chat interface, or automating the data entry nobody wants to do. The goal is always the same: less busywork, more time for the work that actually matters.


Skills

Here’s a brief overview of the technical areas I specialize in:

Computer Vision

ClassificationObject DetectionSegmentation3D Vision3D ReconstructionOpenCV

Data & AI

PyTorchTensorFlowNLPGenerative AI (LLMs)RAG SystemsModel Development/Training/EvaluationProduction DeploymentMLOpsMLflow

Software Engineering

PythonTypeScript/JavaScriptNodeJSExpressJS

Platforms & Delivery

GitPostgreSQLCI/CDDockerKubernetesTerraformGCPAWSAzureScrum/Agile

Professional Experience

KickstartAI

Machine Learning Engineer · 2025 - present

  • Consultancy role in a non-profit helping community efforts and industry partners adopt AI.
  • Projects: retrieval-augmented chatbot (ING); multi-task segmentation and dimension estimation tool for wound-reporting (LUMC).

VDG Engineering & Technology

Self-employed (ZZP) · 2022 - 2025

  • LLM fine-tuning: Mistral-7B on historical Dutch literature (DBNL).
  • RAG search for space standards.
  • News clustering with LLM summarization.
  • LLM-based web-crawl agent for personalized search.

Mainblades Inspections

Lead Engineer, Vision & AI · 2018 - 2025

  • Led development and business implementation of AI solutions; acted as multi-disciplinary system architect reducing aircraft inspection time from eight hours to 20 minutes.
  • Built partnerships with Delta Air Lines, Blue Origin, Lufthansa, and KLM.
  • Highlights: PyTorch damage-detection models; in-house ML stack (Airflow, Docker, Seldon Core); backend contributions (NodeJS/ExpressJS); scalable architecture on GCP/Kubernetes.

Education

Delft University of Technology

MSc. Systems and Control (cum laude) · 2016 - 2019

Thesis: “Classification of Damages on Aircraft Inspection Images Using CNNs”.

ThesisThesis (PDF)Literature Survey (PDF)

Delft University of Technology

BSc. Aerospace Engineering (cum laude) · 2012 - 2016


Contact

[email protected]


Resume

CV (PDF)