Tanguy Pauwels · Freelance ML Engineer
From model
to product.
I build ML systems to extract data from documents, segment images and integrate models into applications.
Computer Vision / Document AI / ML Engineering
01 / Selected projects
Extraction, vision, product integration.
Systems built with imperfect data, limited RAM and review requirements. Each project details measured results and known limitations.
Insureflow
I designed the coverage-table extraction pipeline and PDF review interface. Rewriting the application core with Rust and Python reduced RAM usage by roughly 70%.
Vehicle segmentation
I adapted BiRefNet to client images and built the annotation-to-inference pipeline. Visually observed defects: roughly 50% before, below 25% after.
02 / What I work on
From data to inference.
Document AI
Extract tables and coverage data from heterogeneous PDFs, normalize domain terms and check values against the source document.
Computer Vision
Annotate client images, adapt segmentation models and compare defects before and after fine-tuning.
ML Engineering
Prepare datasets, orchestrate models and integrate inference into an application while controlling memory usage.
03 / ML education
I also explain Machine Learning.
I create animations about Machine Learning, Deep Learning and their mathematics for Instagram and YouTube. Pauwels_ml has over 5,000 Instagram followers. Interactive technical articles are in preparation.
From automotive mechanic to ML Engineer, I kept the diagnostic approach: observe errors, test a hypothesis, measure and iterate.
My backgroundAvailable for freelance projects
Need to improve an ML prototype?
Describe the errors you observe, the data available and your deployment constraints. We can define the next steps and the results to measure.