# Mahesh Ranaweera (ma-he-sh) > Source: https://ma-he-sh.dev/#/ Toronto-based researcher and engineer, bridging simulation and reality. Roles: Senior Software Engineer, PhD Candidate, AI Researcher, Designer ## About I'm **ma-he-sh**, a Toronto-based senior software engineer and PhD candidate in Artificial Intelligence at Ontario Tech University. My research explores sim2real transfer learning, reinforcement learning, and domain randomization to bridge virtual simulations and physical robotics. My engineering background spans full-stack development, scalable backend systems, cloud infrastructure, and automation. Alongside engineering, I experiment with deep learning, LLMs, and robot policies that move from simulation into the real world. When I'm not writing code or reading papers, I'm behind a camera capturing moments, designing 3D prints, or tinkering with electronics. - GitHub: https://github.com/ma-he-sh - LinkedIn: https://linkedin.com/in/maheshranaweera - ORCID: https://orcid.org/0000-0001-7198-5781 - Unsplash: https://unsplash.com/@mahesh_ranaweera ## Experience ### Senior Software Engineer Vocalmeet · July 2018 – present Build and maintain full-stack web applications, design backend services and APIs, and manage cloud infrastructure on AWS and Kubernetes. Develop AI-assisted features and integrate machine learning into production applications. Focus on scalable, reliable systems, accessible interfaces, and automation. ### Communication and Media Coordinator DreBu · Toronto Youth Peer Support · June – Sept 2017 Supported the youth support network through its website, communications, visual materials, and first annual networking conference. ### Website and Graphic Designer DreBu · June 2016 – Jan 2017 Designed an accessible youth support website, developed Android and iOS applications, and created visual identities and outreach materials. ### Robotics Team Lead and Mentor L’Amoreaux C.I. · 2013 – 2014 Led the robotics team in designing and building robots for FIRST and VEX competitions, mentored junior team members, and participated in the FIRST World Championship in Missouri. ## Education ### PhD candidate, Artificial Intelligence Ontario Tech University · 2024 – present Sim2Real reinforcement learning. All academic requirements completed; dissertation research in progress. ### B.Eng., Software Engineering Ontario Tech University · 2014 – 2018 Graduated with distinction. ## Publications ### Bridging the Reality Gap Between Virtual and Physical Environments Through Reinforcement Learning - Venue: IEEE Access (Journal) - Date: March 2023 - Paper: https://doi.org/10.1109/ACCESS.2023.3249572 - Abstract: https://ma-he-sh.dev/docs/publications/2023_paper_ieee_access.md Techniques to train RL models with domain randomization and induced noise to transfer policies from simulation to a physical 3-DOF Stewart platform. ### Evaluation of Techniques for Sim2Real Reinforcement Learning - Venue: The International FLAIRS Conference Proceedings (Conference) - Date: May 2023 - Paper: https://journals.flvc.org/FLAIRS/article/view/133317 - Abstract: https://ma-he-sh.dev/docs/publications/2023_paper_flairs.md Evaluating Q-Learning and Actor-Critic algorithms with domain randomization to bridge the reality gap between simulated and physical environments. ### Bridging Reality Gap Between Virtual and Physical Robot through Domain Randomization and Induced Noise - Venue: Canadian Artificial Intelligence Association (Conference) - Date: May 2022 - Paper: https://caiac.pubpub.org/pub/kzx3gl4e/release/1 - Abstract: https://ma-he-sh.dev/docs/publications/2022_paper_canadian_ai.md Investigating domain randomization and induced noise techniques using deep Q-learning and Actor-Critic methods for sim2real transfer. ### Virtual to Real-World Transfer Learning: A Systematic Review - Venue: MDPI Electronics (Journal) - Date: June 2021 - Paper: https://www.mdpi.com/2079-9292/10/12/1491 - Abstract: https://ma-he-sh.dev/docs/publications/2021_paper_systematic_review.md A systematic review covering transfer learning definitions, recent research, current status, and how it bridges the virtual-to-real-world gap. ## Projects ### PhD Research (active) Bridging the reality gap between simulated and physical environments through reinforcement learning and domain randomization. - **Hexapod Robot**: Working on hexapod robot gait adaptation using Blender simulation. (2024 - current · https://www.youtube.com/embed/ba0NmEVTj_o) ### Sim2Real Transfer (active) Developing techniques to transfer robot policies trained in simulation to real-world hardware. - **MaSC Research**: Master of Science degree research: Sim2Real transfer using Godot. (2024 · https://github.com/ma-he-sh/GodotSim2RealResearch) ### Personal Projects (active) Building full-stack applications for media streaming and interactive learning platforms. - **LumaTerm**: A terminal for developers. (2026 · https://lumaterm.aathma.dev · https://github.com/aathma-dev) - **Syngage**: A platform for students and instructors for active in-class coding. (2023 · https://syngage.io) ### Academic & hackathon projects (completed) Earlier explorations in computer vision, augmented reality, and distributed computing. - **Violence and Assault Recognition**: Lead programmer on a deep neural network platform for detecting violent activity in live and archived video. Second-place capstone project at Ontario Tech University. (2018) - **GeoTourist**: An augmented reality web and mobile application for location guidance, built at QHacks. Contributed the front end and augmented preview. (2018) - **Big Data Benchmarking Platform**: An academic web platform for benchmarking Apache big data frameworks, running test datasets on remote cloud servers, and monitoring execution in real time. (2017) ## Tools & Technologies - **Languages:** Python, TypeScript, JavaScript, C++, Java, Go, C, PHP, Bash - **Backend & Server:** Node.js, Express, Django, REST APIs, GraphQL, Microservices, Flask, Apache Spark - **Cloud & DevOps:** AWS, EC2, S3, Lambda, CloudFormation, Docker, Kubernetes, CI/CD, Amazon EKS, Google Cloud Engine, Git - **Databases:** PostgreSQL, MySQL, MongoDB, Redis, DynamoDB, Elasticsearch, RethinkDB - **AI & Deep Learning:** PyTorch, TensorFlow, Deep Learning, LLMs, Model Training, GPU Computing, CUDA, Keras - **Machine Learning:** Machine Learning, Computer Vision, Reinforcement Learning, OpenCV, Hugging Face - **Robotics:** ROS, Gazebo, Sim2Real, 3D Printing, Sensor Fusion ## Documentation Guides and references: https://ma-he-sh.dev/docs.md ## Photography Portfolio: https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera - https://unsplash.com/@mahesh_ranaweera