My Story
Growing up, my backdrop kept changing —
From the lively streets of Hong Kong to the iconic sights of London. In Toronto, while most were chasing stock prices, I was connecting dots between economics and linguistics at the University of Toronto.
As for Montreal? That's where I really delved deep, making my mark in the realm of AI at McGill. And while academia shaped my perspective, I've taken strides in the corporate world, which allowed me to put my fingerprints on a few patents.
​
SWIPE TO READ MORE :)
My Journey
AI/ NLP Research Scientist
@ GenAIz
Montreal, QC
Sept 2019 - Jul 2023
AI/ NLP Researcher
@ Thales Canada
Montreal, QC
Jul 2018 - Jul 2019
Systems and Methods for Identifying Documents and References (Patent Pending)
As a primary co-inventor, I employed advanced techniques including dependency parsing, regex processing, and open information extraction to develop a state-of-the-art reference annotator. Due to confidentiality agreements, specifics beyond the patent name cannot be disclosed. I invite you to view my portfolio where I showcase other detailed, yet hypothetical and fictional, case studies illustrating my expertise. ​
US Patent Application No.: US 63/399,103
Submission Date: August 28th, 2022
Knowledge-based Recommendation and Insights Engine for Innovation (Patent Pending)
​
As the primary co-author, I pioneered an NLU-driven recommendation system that harnesses the power of knowledge graphs and co-occurrence techniques. This innovation is currently under patent review.
​
US Patent Application No.: US 63/138,633
Submission Date: January 25th, 2021
Open Information Extraction (OIE) System Development
-
Crafted a state-of-the-art OIE system to distill knowledge triples from scientific journals. This was achieved using advanced dependency parsing, rule-based techniques, and syntactic transformations, making it proficient in both English and French.
​
Text Summarization Tool
-
Pioneered a deep learning-driven tool that employs TensorFlow, BERT embeddings, and sequence-to-sequence models to transform extensive texts into succinct, coherent summaries.
​
Semantic Search Research & Implementation
-
Conducted exhaustive research into semantic search paradigms, leading to the development of a cutting-edge semantic search software, optimized for scale and handling over 2 million indexed articles.
​
Image-to-Text Research
-
Investigated contemporary methodologies in Optical Character Recognition (OCR) and deep learning frameworks for converting images into actionable text, paving the way for multimodal data processing.
​
Ontology-based Intent Recognition
-
Led research initiatives focusing on ontology-driven intent recognition, facilitating nuanced and context-aware user intent deciphering in NLU systems.
Database-to-Text Generation​
-
Spearheaded the development of a sophisticated database-to-text generation software, integrating LSTM networks and advanced Bayesian Inference methodologies for optimal data representation and accuracy.
​
Data Analysis & Querying
-
Executed in-depth data analysis on extensive corpora, employing advanced querying techniques using NoSQL and MongoDB, ensuring streamlined data extraction and efficient processing.
Literature Analysis on Text Processing
-
Conducted comprehensive reviews of contemporary literature focused on cutting-edge text processing algorithms. This encompassed a wide array of techniques, spanning Machine Learning models, Neural Network architectures, and their myriad applications.
​
Prototyping & Debugging
-
Conceptualized and brought to life output prototypes, meticulously validated results, and performed debugging. Utilized an array of programming languages and frameworks, including Python, Java, and Theano, to ensure seamless system operations and top-tier software performance.
​
General NLU Projects
-
Delved deep into various NLU tasks, with a keen emphasis on entity recognition, relationship extraction, and semantic role labeling, ensuring comprehensive understanding and manipulation of natural language data.
​
Internal Knowledge Sharing
-
Championed the ethos of continuous learning by hosting knowledge-centric seminars. These sessions were dedicated to elucidating intricate Neural Networks, Machine Learning architectures, and showcasing the prowess of Bi-directional LSTM, Transformer models, and hands-on applications of BERT.
...
For my detailed CV, click HERE