Research Assistant
Constructor University, Bremen, Germany
Working full-time as a Research Assistant in Prof. Dr. Peter Baumann's research group. Focused on scientific research and the development of high-performance geospatial data systems.

Have a project in mind or just want to say hi? Send me a message and I'll get back to you as soon as possible.
Research Assistant · Software Engineer
Constructor University, Bremen, Germany
Working full-time as a Research Assistant in Prof. Dr. Peter Baumann's research group. Focused on scientific research and the development of high-performance geospatial data systems.
Research · Computer Science · Systems
Softgen, Tbilisi
Engineered Java-based integration solutions to automate interactions with the Georgian Revenue Service. Streamlined data exchange processes and collaborated with cross-functional teams to integrate secure APIs.
Java · Spring Boot · REST APIs · Integration
FINA LLC, Georgia
Developed an internal security system using Raspberry Pi clusters. Administered the company's main web platform and received in-depth training on Java EE backend architectures.
Python · IoT · Linux Ops · NGINX
Constructor University, Bremen, Germany
Thesis grew into GSIP, published at IntelliSys 2026.
Buckswood International, Tbilisi, Georgia
Rati Vardiashvili, Dimitar Mishev, Peter Baumann
Big Earth Data — The full systems paper behind GSIP: adaptive memory, seam-free reconstruction, extensibility, and no-data handling, all backed with experiments. Stress-tested on a 17.8-gigapixel, ~1.8-million-km² mosaic (170 Sentinel-2 scenes) on a laptop — 4 GB VRAM, held to an 8 GB memory budget — classified in under two hours.
doi:10.1080/20964471.2026.2746169
Rati Vardiashvili, Dimitar Mishev & Peter Baumann
Proceedings of the Intelligent Systems Conference 2026 (IntelliSys 2026) — Introduces GSIP, the model-agnostic framework and CLI: deep-learning inference on satellite scenes orders of magnitude larger than available memory, scaling from a budget laptop to a high-end workstation.
doi:10.1007/978-3-032-32729-1_47
My Bachelor's thesis: A high-performance inference pipeline for gigapixel satellite imagery. Solves 'grid artifacts' using seamless reconstruction and dynamic memory management.
PyTorch, Rasterio / GDAL, Hydra, GTK4, multiprocessing
A robust FUSE filesystem that mounts iCloud Drive on Linux. Features optimistic syncing, sparse caching for large files, and conflict resolution.
Python, FUSE / pyfuse3, SQLite, GTK3