ML & Infrastructure Engineer with 4+ years of experience designing production ML systems, observability platforms, and scalable backend infrastructure. Lead cross-functional teams, drive architecture decisions at scale, and own end-to-end delivery, from research prototypes to fully automated production deployments.
Currently: ML & Infrastructure Engineer at MDPI, Basel
Years Exp.
Companies
Repos
Products
Work Experience
Selected Projects
Observability Platform
Next-gen monitoring stack with OpenTelemetry, ClickHouse, and SigNoz replacing Grafana-Prometheus.
Infrastructure Migration
End-to-end migration: Airflow → self-hosted Prefect, Poetry → UV, Loguru → Logly/React, ETL with Polars.
API Server Layer
Unified backend for all internal and external services, built from scratch with FastAPI, PostgreSQL, S3.
Clinical AI Platform
Full-stack platform integrating Harvard Medical School AI research into hospital workflows.
Anomaly Detection Pipeline
Industrial ML pipeline (Autoencoder + Siamese) for motor quality inspection at Maxon Group.
Cancer Organoid Tracker
Deep-learning framework (YOLOv5, DeepSORT) for cancer organoid detection, reducing annotation time by 50%.
Education
Master Thesis in Computer Science
MIT — Landmark-Based Co-Registration of Coronary CT and Intravascular Images
2024 — 2024
Master of Computer Science & Life Science
EPFL — Swiss Federal Institute of Technology
2021 — 2024
Bachelor of Biotechnology
MSU — Lomonosov Moscow State University
2017 — 2021
Blog
Building Observable ML Pipelines
How we replaced Grafana-Prometheus with OpenTelemetry + ClickHouse for ML workloads.
From Airflow to Prefect: A Migration Story
Lessons learned migrating orchestration at scale.
Infrastructure as Code for ML Teams
Practical patterns for CI/CD, containerization, and reproducible deployments.
Connect
Feel free to contact me at mariia.erem [at] gmail [dot] com