Hi, I’m Prateek Grover.
I recently completed my Master’s in Computer Science at KU Leuven.
I work on computer vision: segmentation, tracking, and multimodal sensing — across research labs where I’ve run the experiments as well as built the models.
I also work on the engineering around them: training pipelines, cloud and production infrastructure, and interfaces built for people who don’t write code.
Education
Experience
Platform & DevOps Engineer
Atlas Copco Group
Oct 2025 – Present · Antwerp, Belgium
Built observability for a distributed industrial IoT platform (dashboards, alerting, cross-tenant metrics) and set up Kubernetes clusters with GitOps delivery now running databases and services in development. All managed as infrastructure as code.
View details →
Machine Learning Research Intern
Life Sciences Dept., imec
Oct 2024 – Jun 2025 · Leuven, Belgium
Built a platform letting biologists run and fine-tune segmentation models on their own data without writing code. Ran the wet-lab experiments behind it, producing a novel dataset pairing fluorescence microscopy with impedance recordings.
View details →
Software Engineering Intern
Nokia Bell Labs
Jul 2024 – Aug 2024 · Antwerp, Belgium
Designed a microservices system for autonomous monitoring and remediation of broadband network units, and applied hierarchical clustering to infer unknown network topology from connectivity patterns alone.
View details →
Software Engineer
Walmart Global Technology Services India
Jul 2022 – Aug 2023 · Bengaluru, India
Extended Sam's Club's shipment tracking platform across domestic and ocean freight — owning testing infrastructure, event pipeline reliability, production alerting, and a greenfield maritime tracking initiative.
View details →
Summer Research Associate
Center for Computational Biology, Flatiron Institute
May 2022 – Jul 2022 · New York, United States
Trained generative models to synthesise biological imaging data, and built ConvNets for nuclei tracking and cell-cycle prediction — reconstructing full lineages from time-series microscopy, including through cell division.
View details →
Research Intern
Center for Research in Computer Vision, University of Central Florida
Dec 2021 – May 2022 · Florida, United States
Benchmarked segmentation models on a novel mouse embryo dataset, then designed U-Net and Vision Transformer architectures using hybrid contour-distance representations to push accuracy past existing baselines.
View details →
Select Projects
Minimal Interpreter for a Domain-Specific Language
KU Leuven - Programming Languages
Nov 2024 – Dec 2024
A lightweight interpreter for a custom language, written in Racket — arithmetic, first-class functions, type checking, and error handling.
Optimised Delivery Routing with Graph Neural Networks
KU Leuven - Constaint Solving
Oct 2023 – Apr 2024
A GNN-guided solver with clustering to optimise delivery routing, improving route similarity and efficiency across 100+ deliveries per day.
Language Models for Structured Data Prediction
BITS Pilani - Reseach Project
Sep 2021 – Mar 2022
Predicted material properties by parsing chemical nomenclature as a structured language, benchmarking transformer models (BERT, RoBERTa) against graph networks for how well each captured semantic structure.
JP Morgan Quant Research Challenge — Winner
JPMorgan Chase
Aug 2021 – Sep 2021
Won the challenge with a model for long- and short-range time-series forecasting of asset prices, used to construct a diversified portfolio.
Real-Time Pose Estimation and Movement Scoring
BITS Pilani - Research Project
Jun 2021 – Dec 2021
Scored martial-arts technique against expert reference recordings: skeletal keypoints from a live webcam feed, normalised for position and body scale, compared by cosine similarity.
Neuroevolution for Deep Q-Learning
BITS Pilani - Neural Networks
Mar 2021 – Apr 2021
Trained a reinforcement learning agent to play Atari games with a Deep Q-Network, using neuroevolution to search the hyperparameter space rather than tuning by hand.
Corpus Search Engine with Vector Space Ranking
BITS Pilani - Natural Language Processing
Mar 2021 – May 2021
A ranked information retrieval system over a 25k+ document corpus, with a custom in-memory index for low-latency querying and a vector space model using TF-IDF and cosine similarity.