Romi Kuntsman is a data engineer whose work sits where large-scale data systems meet scientific and machine-learning research — from recommendation systems and computer vision to drug discovery and medical AI.
Engineering for research, not just for dashboards
His LinkedIn summary is unusually direct about the goal: building reliable, scalable data pipelines that enable scientific research. The work described across his recent roles is less about generic analytics and more about making complex research data usable, traceable, reproducible and ready for machine-learning workflows.
At GSK, for example, he describes building compliant ETL pipelines that turned genomic and clinical data into ML-ready datasets, structuring data for traceability and controlled access, and connecting DataOps and MLOps tooling across HPC and cloud environments.
A recurring theme in his career is building the layer between raw information and the people or models that need to trust it.
Career
Data Engineer for Medical AI
Current role in Tel Aviv-Yafo, working in the medical-AI field.
Data Engineer for Drug Discovery Research
Compliant ETL for genomic and clinical data, research dataset traceability, controlled access, DataOps/MLOps integration and mentoring researchers on data, AI and programming tools.
Data Engineer for Computer Vision Research
Pipelines for active learning, annotation workflows and dataset generation, plus tooling for preprocessing, deployment support, ingestion, validation and visualization.
Senior Big Data Engineer
Multi-tenant batch and streaming pipelines using Spark and Elasticsearch, with work on performance, cost, modeling, partitioning and orchestration.
Data Engineer for Recommendation Algorithms
Worked on the personalization team, developing recommendation systems based on large-scale data analytics.
Software Engineer
R&D involving browser engines, web crawling, databases, networking, cloud systems and user-interface frameworks.
System Performance Engineer
Software and database performance optimization, high-throughput event handling, real-time web-scale systems and cloud technologies.
Systems, Monitoring & Automation
Several infrastructure roles spanning systems administration, datacenter monitoring, management tooling and automation.
Writing and experiments
Outside his work history, Kuntsman publishes Romi's Tech Blog on Substack. The blog gives the page a more personal side: not a corporate bio, but a place for technical experiments and questions about software, data and AI.
Hangman test for LLM — who wins?
A small experiment using Hangman as a test of consistency in language models — asking whether a model can keep the same hidden word stable across a conversation rather than merely sounding convincing from turn to turn.
Read the post on Substack ↗Education
M.Sc. Computer Science
The Open University of Israel
2018–2023
B.Sc. Computer Science & Life Sciences
The Open University of Israel
2006–2013