Dylan Ler

Dylan Ler is an accomplished data scientist and machine learning engineer with a strong background in AI and deep learning. With extensive experience in developing innovative solutions, Dylan has a proven track record of driving impact in various industries. Passionate about technology, Dylan actively contributes to hackathons and engages in continuous learning.

DYLAN LER

EDUCATION
Georgia Tech                                                                                                                                                                     
Masters in Computer Science - Concentration in Machine Learning (Remote Online/Part Time - Ongoing)
Machine/Deep Learning: TensorFlow, PyTorch, BERT, NLTK, Pandas, Jupyter, Scikit-learn, OpenCV, AWS, Sagemaker, GCP, Docker, Langchain

William Jessup University                                                                                                                                               Class of 2021
Masters in Computer Science - Concentration in Technology Management 
Technical skills: Python, C++, Java, SQL, Git, Matlab, Android Development, Flask, Gradio, Tableau, Matplotlib, Seaborn

University of California, Los Angeles  Class of 2016 BSc. Electrical Engineering
Computer Vision, Computational Photography, Speech Recognition, and System Processing

PROFESSIONAL EXPERIENCE
WebMD () - Internet Brands Inc.  Los Angeles, California
Senior Data Scientist & Machine Learning Engineer   November 2016 - Present
Built a recommendation system for medical forums under WebMD using BERT embeddings and an LSI topic modeling model to increase user engagement and retention.
Developed, tested, & installed an AI tool that understands medical articles semantically, extracts important keywords from articles using PositionRank, and automatically hyperlinks keywords to internal medical articles.
Used NLP techniques and K-means clustering to successfully train and deploy a model that classifies text messages to aid physicians & doctors with patient appointments.
Built a churn prediction model using XGBoost with 98% accuracy to predict when a customer will cancel their subscription to in-house WebMD patient management software.
Built an AI chatbot that can answer questions in natural language about medical information from WebMD’s internal database using Langchain and Gradio.
Built a recommendation engine to recommend email newsletter to WebMD’s email newsletter which retained a more than 90% open rate using deep learning.
Created data visualization dashboards using Tableau to display visitor data on webmd in order to make analytical decisions to optimize the user experience as well as SEO purposes.
Participated in an internal company hackathon to create an AI music therapy tool that auto-generates relaxing music in any desired genre like lo-fi, classical, and hip-hop using an LSTM network.
Skills used: TensorFlow, PyTorch, NLTK, Pandas, Scikit-learn, BERT, Huggingface, Spark, Hadoop,  Langchain
Udacity () Los Angeles, California
Mentor & student for AI  Engineer Nanodegree Program                                                   March 2017- June 2018
Guided 50 students in topics of Deep Learning, Computer Vision, and C++ to build software for a cancer detection project.
Reviewed student code projects in Edge Detection, Image  Classifiers, Data Augmentation, and Feature Engineering
Skills used: Python, C++, Git

INVENTIONS & PROJECTS
Aki Story NFT Project ()                                Global
Software | Blockchain Developer          April 2022-Ongoing
Developed smart contracts and a front-end website for an NFT project that created and fully distributed 5,555 NFTs with user utilities such as a virtual metaverse, educational material, and add-on NFTs on the Ethereum blockchain.
Skills used: Solidity, HTML, CSS, Javascript, Java, Python, Web3.js

UCLA Wireless Health Institute & IEEE    Los Angeles, California
Software | Electrical Engineer                                                                                                                    September 2013-June 2015
Designed and built PRIME—an Android application that combines sensor tracking with medication & survey prompting 
to aid physicians and patients in progressing through the recovery phase.
Won competition creating gesture-based sign language communication devices using Myo band & Arduino (C++).

HIGHLIGHTS
Mentor for hackathons (software programming competitions): - Cal Hacks (UC Berkeley) & HackingEdu (San Mateo)
Author of a personal finance and investment blog to help people manage and practice healthy financial habits.
Hobbies: hiking, implementing deep learning models from research papers, and managing online communities.

Masters in Computer Science - Concentration in Machine Learning (Remote Online/Part Time - Ongoing)

• Machine/Deep Learning: TensorFlow, PyTorch, BERT, NLTK, Pandas, Jupyter, Scikit-learn, OpenCV, AWS, Sagemaker, GCP, Docker, Langchain

William Jessup University                                                                                                                                               Class of 2021

Masters in Computer Science - Concentration in Technology Management 

• Technical skills: Python, C++, Java, SQL, Git, Matlab, Android Development, Flask, Gradio, Tableau, Matplotlib, Seaborn

University of California, Los Angeles  Class of 2016 BSc. Electrical Engineering

• Computer Vision, Computational Photography, Speech Recognition, and System Processing