I'm a software engineer with five years of experience building data-intensive systems. For the past year I have been studying machine learning seriously, with the goal of moving into research.
Currently:
Research & Learning
Self-directed projects I use to build a foundation for machine learning research. All code and results are public, and the results tables only contain numbers from runs I actually made.
Comparison of cost-sensitive weighting, SMOTE and random undersampling against untreated baselines on three medical datasets, using 5-fold stratified cross-validation scored with PR-AUC instead of accuracy.
Committed result: balancing methods mostly improved recall rather than ranking, and made little difference when imbalance was mild.
Reproducible PyTorch training harness for CIFAR-10 with LeNet and a CIFAR-stem ResNet-18: deterministic seeding, cosine learning-rate schedule, per-epoch metrics logged to CSV.
LeNet reaches 80.25% test accuracy in 30 epochs; the ResNet-18 run is in progress.
Text-to-image search over local photo collections using pretrained CLIP embeddings, with an incremental embedding index and a command-line interface for natural-language queries.
Working retrieval tool; index updates without re-embedding unchanged images.
Worked exercises from Software Foundations Volume 1 in the Coq proof assistant: functional programming and proofs by induction, checked by the proof checker.
Ongoing; Basics chapter complete.
Selected Work
Products and platforms I have built as a software engineer.
Navian Platform Full-stack engineer Real estate analytics platform with market intelligence reports, subscription billing, and investment tools for the Swedish market. TypeScript Node.js React PostgreSQL Python 
OceanDrive Full-stack engineer Decentralized storage platform for peer-to-peer file distribution across networked nodes. React Next.js Node.js PostgreSQL 
World Art DEXPO Frontend engineer Digital art marketplace built with React and NestJS, with wallet-based authentication and on-chain asset ownership. React NestJS Web3 What I Work On
Engineering is my profession; machine learning is where I am heading. The two overlap more than they sound.
Five years building production systems: data pipelines, APIs, and analytics platforms used daily by real customers.
Working through the fundamentals in public: classical models, evaluation methodology, and PyTorch training, with code and results on GitHub.
A physics and mathematics background, kept active: probability and statistics, linear algebra, and mechanized proofs in Coq.
Writing
Notes on things I have built and debugged in production.
What I learned from benchmarking cost-sensitive weighting, SMOTE, and undersampling on three medical datasets, with real numbers from cross-validated runs
A software engineer's habits applied to PyTorch experiments: deterministic seeding, metrics as artifacts, and why an interrupted training run proved the point
Learn how to implement robust Stripe payment systems with subscription management, billing cycles, and secure webhook handling for SaaS platforms
