jwnicholas:/about$ whoami
Jing Wei Nicholas Lim
Member of Technical Staff @ OpenAI
Computer Science & Philosophy @ Brown
Hi, looks like you made it down here - I'm Nicholas and welcome to my corner of the internet! Check out some of my projects above and my relevant experiences below. To the right are some of the things I write on in my blog.
I'm a maker - I like building and working with robots as they have a tangible effect on the real world + I also like working with my hands!
My interests are mostly concentrated around the application of AI/ML. In pursuing this, I'm currently working at OpenAI, and previously conducted research in reinforcement learning at the Intelligent Robot Lab @ Brown and in robotics at the Humans To Robots Lab @ Brown.
I want to use my programming knowledge to solve real-world problems - send me an email at jwnicholas99@gmail.com if there is something you want to collaborate on!
Working on AI safety.
Built and launched Postman, Singapore’s whole-of-government secure messaging platform that unifies all government SMS messages under one “gov.sg” sender ID. With >300 million messages sent, citizens can easily identify genuine messages from the government and avoid government impersonation scams.
Led design and implementation of decentralized finance (DeFi) app facilitating payments via Ethereum smart contracts, further integrating with SAP’s Self-Sovereign Identity (SSI) platform. Stepped up as tech lead of SSI during team transition, onboarding and leading 4 new engineers.
Built and deployed end-to-end high-throughput service in Golang processing 1K+ requests/s about Captcha challenges in real-time across tens of millions of users. Engineered scalable ETL pipeline using Airflow DAGs processing millions of data points for business intelligence and fraud detection.
Improved object detection performance by 20% on personal mobility devices compared to previously deployed YOLOv3 model, through automating data extraction/annotation and greatly expanding hyperparameter search space using Bayesian optimization.
Publication: Skill Discovery for Exploration and Planning using Deep Skill Graphs (DSGs), ICML 2020 LifelongML - used DSGs to reduce large continuous MDPs into small discrete MDPs. Built AI training and inference pipelines for hierarchical reinforcement learning (HRL) research. Researched option termination classifiers for image-based observation spaces in HRL
Implemented DataQuality framework to identify missing/malformed data in >20 categories of information in millions of health records.
Publication: A Brain-Computer Interface (BCI) to Detect Responses to Affective Audiovisual Stimuli from Electroencephalogram (EEG), IRC-SET 2017. Architected an AI-powered BCI to control a robot wheelchair using EEG and motor imagery