Ronen Abravanel
Senior Software Engineer & Computer Vision Specialist
15+ years of expertise in AR/VR spatial computing, high-performance C++ development, and SLAM infrastructure at Meta and Magic Leap, with a strong academic foundation in physics and published research.
Skills
At a Glance
15+
Years of Software Engineering Experience
100-node
Parallel Computing Cluster (MPI/HPC)
2
Top-Tier Tech Companies (Meta, Magic Leap)
1
Published Research in Physical Review D
Work Experience
Senior Software Engineer, Computer Vision
Meta (Reality Labs) · July 2020 – Present
Leading the development of a distributed cloud-based localization infrastructure for next-generation smart glasses. Implemented high-performance C++ Thrift services and ZippyDB integration for large-scale spatial data persistence, enabling cross-device world-locking and long-term spatial anchoring. Built core on-device localization for Meta Quest with performance-critical optimizations for mobile chipsets, and developed comprehensive benchmarking pipelines using Hive analytics on large-scale telemetry data.
Senior Software & Computer Vision Engineer
Magic Leap · 2017 – 2020
Led the research pipeline for Magic Leap 2 head-tracking, driving SLAM, Pose Estimation, and Stereo Vision algorithms from prototype to production. Designed and built a full-stack automated system for data analysis and 3D visualization using Vue.js, Three.js, and Plotly, enabling real-time debugging and visualization of complex spatial data for the research team.
Software & Algorithms Developer
Visionmap (RAFAEL) · 2014 – 2017
Developed complex photogrammetry and image processing algorithms for defense-grade geospatial applications using high-resolution aerial imagery. Implemented computationally efficient, scalable C++11 solutions for large-scale geospatial data processing.
Research Assistant
Technion – Department of Physics · 2011 – 2014
Developed large-scale Monte Carlo simulations to study superconductor phase transitions on a 100-node parallel computing cluster using C++ and MPI. Performed complex data analysis and visualization of statistical mechanics results using Python.
Software Developer
IDF Military Intelligence – Unit 8200 · 2004 – 2008
Contributed to a project awarded the Israel Defense Prize in 2008. Led software R&D projects focusing on high-performance C++ and Win32 API development, with full-stack expertise across Python, C#, Perl, and Shell scripting.
Key Projects
Distributed Cloud Localization Infrastructure
Architected and implemented a scalable cloud-based localization system for Meta Quest and next-generation smart glasses, leveraging Thrift services, Meta Ent/Node API, and ZippyDB for persistent spatial data. Developed novel sub-component metrics and multi-session evaluation systems to measure accuracy and long-term map stability.
Magic Leap 2 Head-Tracking Research Pipeline
Led development of core SLAM and Pose Estimation algorithms for monochrome camera inputs. Built a full-stack visualization and analysis system using Vue.js, Three.js, and Plotly for real-time algorithm debugging and performance assessment.
High-Performance Monte Carlo Simulation
Designed and executed large-scale simulations studying superconductor phase transitions on a 100-node parallel cluster using C++ and MPI, analyzing complex statistical mechanics data with Python visualization.
Achievements
- Israel Defense Prize (2008) for significant technical achievement in software development
- Published research in Physical Review D on Casimir interactions and the Babinet Principle (MIT, 2011)
- Led Python Workshop for Physics students at Technion, teaching scientific computing and data analysis