William Ma
Greater Vancouver Metropolitan Area Canada
Summary
🤩
Rockstar🎓
Top SchoolWilliam Ma is an AI engineer with 12 years' experience building research-driven systems across reinforcement learning, computer vision, NLP and probabilistic reasoning. He contributed core NLP and rule-engine improvements to the open-source OpenCog AGI framework and co-led the open-sourcing of SenseAct, the first RL toolkit for physical robots while on Kindred’s AI research team. His background spans applied R&D—shipping AR/vision solutions for global brands—to production ML systems for marketing automation and customer support. Trained in computer vision and medical imaging at Simon Fraser University, he combines deep research pedigree with pragmatic engineering for embodied and language-aware AI. Notably, his work on backward-chaining, R2L parsing rules, and microplanning shows a rare blend of symbolic reasoning and practical parser/engine engineering.
12 years of coding experience
7 years of employment as a software developer
Bachelor of Science, Mathematics, Bachelor of Science, Mathematics at The University of British ColumbiaMaster of Science, Computer Science, Computer Vision, Medical Imaging, Master of Science, Computer Science, Computer Vision, Medical Imaging at Simon Fraser University
Github Skills (13)
scheme10
nlp10
text-parsing10
parsing10
atomic10
rule-engine10
atomics10
lisp9
graph-database9
pattern-matching9
knowledge-representation9
logic-programming8
query-engine7
Programming languages (6)
Github contributions (5)
opencog/opencog
May 2014 - Nov 2015
A framework for integrated Artificial Intelligence & Artificial General Intelligence (AGI)
Role in this project:
Back-end Developer
Back-end DeveloperContributions:473 commits, 156 PRs, 85 pushes in 1 year 6 months
Contributions summary:William's primary contributions centered around enhancements and modifications within the natural language processing (NLP) components of the OpenCog project. The commits demonstrate work on rules within the NLP pipeline, specifically Relex-to-Logic (R2L) rules and microplanning. The user implemented new R2L rules and worked on the design of the parser and associated helper functions that generate code in scheme. They also addressed a related issue concerning graph based representations.
unsupervised-learning-algorithmsroboticsnatural-language-understandingartificial-general-intelligenceintegrated
opencog/atomspace
May 2015 - Nov 2015
The OpenCog (hyper-)graph database and graph rewriting system
Role in this project:Back-end Developer & Rule Engine Specialist
Back-end Developer & Rule Engine SpecialistContributions:199 commits, 82 PRs, 43 pushes in 6 months
Contributions summary:William primarily focused on enhancing and debugging the rule engine component within the atomspace repository. They made changes to the backward chainer, including adding step limits, implementing random target selection, and handling cases with no premises. The user also implemented a focus set, a sub-atom unification solution, and refactored code to incorporate Target objects and handle variable declarations. These changes suggest an effort to improve the rule engine's functionality and performance.
logic-programmingopencogknowledge-baseedgedbknowledge-graph
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.