
adian
Spatial Computing Engineer, 3D Computer Vision
Habilidades

Conheça meus serviços


Experiência profissional
Senior 3D Vision Engineer · Lead
Spatial Dynamics Labs • Período integral
Jun 2022 - Present • 3 yrs 11 mos
Leading a cross-functional team of 8 engineers building next-generation spatial reconstruction pipelines for industrial digital twin platforms, serving Fortune 500 clients across oil & gas, aerospace, and manufacturing. Architected a real-time 3D Gaussian Splatting renderer achieving 120 fps on consumer hardware, reducing scene loading time by 78%. Led development of a hybrid NeRF–SDF pipeline for sub-millimetre precision industrial metrology deployed across 14 offshore oil rigs. Designed LiDAR–camera fusion system processing 2M points/sec for autonomous facility inspection drones. Built BIM-integrated digital twin platform connecting IFC models to live sensor data for 3 major infrastructure projects ($2B+ combined value). Filed 3 patents on real-time mesh simplification using octree-guided SDF sampling.
Computer Vision Engineer · Mid-Senior
ReconTech Systems • Período integral
Mar 2020 - Dec 2021 • 1 yr 9 mos
Developed photogrammetry and multi-view stereo pipelines for the construction industry, enabling automated progress monitoring and as-built documentation from drone imagery at scale. Built an automated SfM pipeline using COLMAP + custom bundle adjustment, reducing processing time from 6 hours to 22 minutes. Implemented TSDF volumetric fusion for seamless multi-session scan alignment with <2mm error across 500m² areas. Developed point cloud semantic segmentation model (PointNet++ based) achieving 94.3% mIoU for construction material classification. Created SDK used by 200+ enterprise clients for integrating 3D outputs into existing BIM workflows.
Research & XR Engineer · Junior–Mid
XR Innovations Studio • Período integral
Sep 2018 - Feb 2020 • 1 yr 5 mos
Focused on AR/VR spatial computing research, prototyping novel 3D representation methods for interactive mixed reality experiences in game development and location-based entertainment. Prototyped one of the first real-time NeRF-based scene streaming systems for HoloLens 2, presented at CVPR 2020 workshop. Built Unity plugin for real-time SDF-based soft body physics achieving 2× performance vs. traditional mesh methods. Developed voxel-based scene understanding pipeline for AR occlusion, deployed in 5 location-based VR attractions. Co-authored 4 peer-reviewed papers on neural implicit representations and real-time rendering.