Siddharth Gupta
Geospatial Software Engineer · Python, 3D Point Clouds, LiDAR Simulation, Sensor FusionSUMMARY
Geospatial and software engineer with 7 years of professional experience across geospatial data processing, 3D point-cloud analysis, LiDAR simulation and application quality assurance.
Experienced in developing Python-based geospatial workflows, building sensor simulation tools, and translating research concepts into practical software. Contributed to peer-reviewed research at IIT Kanpur and IIRS Dehradun. Also developed open-source tools, including four published QGIS plugins with over 10,000 downloads and QuickPointForge, a tool for generating LiDAR-style point clouds from 3D Gaussian Splat scenes. Combines geospatial domain knowledge with software development, testing and problem-solving skills to build reliable data-processing tools and engineering workflows.
CORE SKILLS
3D PERCEPTION & POINT CLOUDS
Point-cloud generation, labelling and ML-based classification; Gaussian Splatting → point-cloud conversion via spherical binning; LiDAR–camera sensor fusion for 3D reconstruction; mesh/point-cloud QA and inspection.
SENSOR SIMULATION & SYNTHETIC DATA
Custom LiDAR/camera sensor simulation engine in Python (octree spatial indexing, ray casting); Blender-based configurable LiDAR/camera sensor object; synthetic dataset generation and SOPs for ML training data.
SPATIAL MAPPING & GEOSPATIAL DATA
Photogrammetry, remote sensing, multi-source (LiDAR + imagery) data integration; QGIS, ArcGIS; scalable multiprocessing raster/point-cloud pipelines.
QA, TEST AUTOMATION & RELIABILITY
Functional/regression testing, browser-based test automation (Selenium, Cypress), defect management (JIRA), deployment validation, OWASP-oriented security testing.
EXPERIENCE
Independent GIS & Software Consultant
- Delivered geospatial data engineering and processing workflows across QGIS, LiDAR and remote sensing for consulting clients, handling large-scale spatial datasets.
- Developed scalable, multiprocessing Python pipelines for efficient geospatial data processing.
- Published and maintained 4 QGIS plugins (QuickMapCine, QuickMapCompare, QuickMapGif, QuickMapLink) on the QGIS Plugin Repository, reaching 10,000+ cumulative downloads within a month and 5-star ratings.
Junior Research Fellow
MAR 2021 — APR 2023- Built a new version of Limulator, the lab’s synthetic LiDAR simulator, in Python using octree spatial indexing and ray casting, then rebuilt it as a Blender plugin with a configurable LiDAR/camera sensor object that outputs synthetic point clouds and images.
- Produced synthetic datasets used to train ML models for point-cloud classification, and authored SOPs standardizing labelled dataset preparation.
- Co-authored a peer-reviewed paper analyzing the role of simulated LiDAR data in training 3D deep learning models (see Publications).
Consultant
FEB 2021 — MAY 2026- Owned functional and regression testing for multiple web applications, including full user-workflow and post-deployment validation, and built browser-based test automation with Selenium and Cypress.
- Established test planning and QA documentation practices and applied OWASP-oriented security testing, coordinating remotely with developers to verify fixes and validate releases.
Senior Research Fellow
DEC 2020 — JAN 2021- Trained end users on a GIS portal and sample-selection application through detailed documentation and live interactive sessions.
Test Engineering Analyst
AUG 2015 — DEC 2016- Owned test planning, preparation and execution through closure; conducted requirement analysis for new and modified products.
SELECTED PROJECTS
QuickPointForge
2026- Built a tool that simulates LiDAR point clouds directly from 3D Gaussian Splats, binning splats by sensor beam angle to keep the process lighter than ray casting.
- Built a desktop UI with a dual-panel viewport: the source Gaussian splat with the sensor pose shown as a gizmo, and the resulting simulated LiDAR scan.
Pixly
2025 — 26- Designed and developed a distributed application comprising a REST API backend, a host administration desktop app, and an Android app for guests and admins.
- Built a self-hosted face-recognition pipeline using InsightFace, ONNX Runtime and DBSCAN clustering, deployed as a separate background worker for asynchronous photo processing.
- Separated computationally intensive image processing from the core application so photo uploads and user interactions never block.
EDUCATION
M.Tech, Remote Sensing & GIS
2020- CGPA 8.20 · GATE Score 508 (2021), All-India Rank 171
- Thesis: Framework for multi-source data integration (LiDAR + camera/drone imagery) for 3D documentation of heritage sites, using fractal analysis to reconstruct damaged areas.
B.E., Civil Engineering
2015- CGPA 8.45 · GATE Score 490 (2018)
PUBLICATIONS
Tiwari, P. S., Pande, H., Gupta, S., Grover, C., Semwal, E., & Agarwal, S. (2023). “Damage Detection and Virtual Reconstruction of Built Heritage: An Approach Using High-Resolution Range and Intensity Data.” Journal of the Indian Society of Remote Sensing, 51(4), 787–798. DOI ↗
Lohani, B., Khan, P., Kumar, V., & Gupta, S. (2024). “Role of Simulated Lidar Data for Training 3D Deep Learning Models: An Exhaustive Analysis.” Journal of the Indian Society of Remote Sensing, 52(9), 2003–2019. DOI ↗