As the Founding Machine Learning Engineer at Minimist, I lead the design, development, and deployment of end-to-end machine learning solutions, leveraging foundational models, Bayesian statistics, and the Python data science stack. My role involves creating intelligent systems that streamline workflows and save time for our customers. Minimist is transforming the second-hand market by empowering stores to digitize their inventory quickly with our innovative scanner app.

Led the creation of Loupe, a new product, used by design brands such as Hay, Gubi and Flos• Automated 70% of manual product matching through a combination of custom PyTorch modelsserved on Vertex AI through TorchServe and foundational Visual Language Models• Reduced our scraping cost by 60% after migrating from Cloud Run to Kubernetes• Architected a data workflow orchestration using Dataflow, Dataproc and Google Cloud Workflows• Hired and supervised an ex-Google Master Thesis Student• Fully implemented a Web App from scratch using FastAPI and React

Tackled the challenge of Entity Resolution in e-commerce using PyTorch and transformers• Designed and deployed a novel Contrastive Learning approach, improving on methods from theliterature by 9.7% F1 score on average on data scarce scenarios• Analyzed large datasets using PySpark, pandas, and numpy for model training and evaluation

Developed an intelligent agent capable of extracting tabular data from scanned documents• Enabled users to import projects in seconds instead of manually typing them in days or even weeks• Integrated multiple OCR technologies including Google Cloud Vision, Azure AI and Tesseract• Implemented a robust, well tested Java-based solutions to handle various document formats

Led a team of 4 developers, mentoring and fostering the growth of my colleagues• Architected a modern tech stack using Spring Boot and Angular• Managed the full software development lifecycle from initial design to deployment and maintenance• Enabled the growth of the company’s Annual Recurring Revenue (ARR) by a factor of 4x