Wenbo (Vito) Zhu
Founding AI Engineer at Opus Clip- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
0
/5.0 / Based on 0 ratingsFilter reviews by:
Credentials
-
Sliver Medal - ASHRAE - Great Energy Predictor III (rank 62/3614, top 2%)
KaggleJan, 2020- Sep, 2024 -
Managing Big Data with MySQL
CourseraSep, 2018- Sep, 2024 -
Data Visualization and Communication with Tableau
CourseraJun, 2018- Sep, 2024 -
Convolutional Neural Networks
CourseraMay, 2018- Sep, 2024 -
Deep Learning Specialization
CourseraMay, 2018- Sep, 2024 -
Mastering Data Analysis in Excel
CourseraMay, 2018- Sep, 2024 -
Sequence Models
CourseraMay, 2018- Sep, 2024 -
Business Metrics for Data-Driven Companies
CourseraApr, 2018- Sep, 2024 -
Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
CourseraApr, 2018- Sep, 2024 -
Neural Networks and Deep Learning
CourseraApr, 2018- Sep, 2024 -
Structuring Machine Learning Projects
CourseraApr, 2018- Sep, 2024
Experience
-
Opus Clip
-
United States
-
Software Development
-
1 - 100 Employee
-
Founding AI Engineer
-
Mar 2022 - Present
Lead AI of Opus Clip, No.1 AI Video Clipping Tool in the Market, reaches $1M ARR in 1-month, with 400K+ users and 30M+ clips generated. With a mission to Unleash Creativity and Authenticity with AI, we are building Opus Clip to democratize video content creation, making it accessible to anyone with a story to tell or a passion to share—not just those with expertise in video editing. Backed by Lightspeed, DCM, and several other investors. Lead AI of Opus Clip, No.1 AI Video Clipping Tool in the Market, reaches $1M ARR in 1-month, with 400K+ users and 30M+ clips generated. With a mission to Unleash Creativity and Authenticity with AI, we are building Opus Clip to democratize video content creation, making it accessible to anyone with a story to tell or a passion to share—not just those with expertise in video editing. Backed by Lightspeed, DCM, and several other investors.
-
-
-
Harbin Institute of Technology(Shenzhen)
-
China
-
Higher Education
-
1 - 100 Employee
-
External AI Expert
-
Aug 2023 - Present
Partenership and external AI expert of the academician workstation of MEI HONGYUAN in south China, Harbin Institute of Technology(HIT) Partenership and external AI expert of the academician workstation of MEI HONGYUAN in south China, Harbin Institute of Technology(HIT)
-
-
-
ByteDance
-
China
-
Software Development
-
700 & Above Employee
-
Senior Machine Learning Engineer
-
Jul 2020 - Mar 2022
Founding team of Gauthmath (app solving math problems using AI), reaches 6M MAU and 100M downloads in 1-year # Owner of the geometry problem solving service, the world's 1st online app service which can solve geometry math problems by AI # Initial eng team member of Gauthmath app (MAU >600w, 0 to 1) # Owner of the document understanding service, including OCR(text, formula) and diagram understanding # Responsible for building content-based image retrieval system Founding team of Gauthmath (app solving math problems using AI), reaches 6M MAU and 100M downloads in 1-year # Owner of the geometry problem solving service, the world's 1st online app service which can solve geometry math problems by AI # Initial eng team member of Gauthmath app (MAU >600w, 0 to 1) # Owner of the document understanding service, including OCR(text, formula) and diagram understanding # Responsible for building content-based image retrieval system
-
-
-
云从科技 Cloudwalk
-
China
-
IT Services and IT Consulting
-
1 - 100 Employee
-
Machine Learning Engineer
-
Jul 2019 - Jul 2020
Team member of Structured Data Research Institute at Cloudwalk #1 Designed a billon-scale, incremental face clustering algorithm based on K-means and birch with recall greater than 97% #2 Developed an engine for similarity search and clustering of vectors (qps > 2000) including HNSW and Bruteforce index #3 Add auto-tuning method for clustering parameters and features transforming method for Face Clustering Product #4 Awarded "Best Team Award" in Q3 as a core member for designing and deploying our Face Clustering Product in four provinces in both finance and security field Cloudwalk is one of the Top 4 AI Unicorn company in China Show less
-
-
-
Prosper Marketplace
-
United States
-
Financial Services
-
300 - 400 Employee
-
Data Scientist
-
Aug 2018 - May 2019
#1 Query dataset of more than 3 million samples from customer loan, demographic and third-party credit database within 6-months window in P2P industry #2 Conducted data preprocessing and feature selection on 90+ variables with GBDT, SMOTE, WOE encoding, etc #3 Utilized Undersampling method to deal with imbalanced dataset and built a data pipeline for feature engineering process #4 Built Xgboost and Logistic models to identify whether the customer will return for another loan with the AUC of 0.9 and built Generative models to predict the probability of their returning Prosper is America's first marketplace lending platform with over $10 billion in funded loans. Show less
-
-
-
Paradigm
-
United Kingdom
-
Defense and Space Manufacturing
-
1 - 100 Employee
-
Data Scientist
-
Aug 2018 - Dec 2018
#1 Scraped more than 100,000 articles with Newsapi and web resources and marked price change >2SD with Coinapi #2 Extracted features including description, publisher, tf-idf, sentiment score from articles with nltk, Google NLP API #3 Built semi-supervised models including TSVM and Graph-based models to rank the importance of the articles with accuracy of 70% #1 Scraped more than 100,000 articles with Newsapi and web resources and marked price change >2SD with Coinapi #2 Extracted features including description, publisher, tf-idf, sentiment score from articles with nltk, Google NLP API #3 Built semi-supervised models including TSVM and Graph-based models to rank the importance of the articles with accuracy of 70%
-
-
-
Beihang University
-
Software Development
-
700 & Above Employee
-
Research Assistant
-
Sep 2016 - Jun 2018
Mainly designed and conducted two experiment project in Human Factors: #1 The Influence of Different Chinese Tones on Human Cognitive Load based on Human-Computer Interaction (1) Selected the topic based on literature review and weekly discussions and set different time limit, tones of Chinese, sex and devices (iPhone or iWatch) as independent variables (2) Designed the flow of experiment utilizing E-prime 2.0 and finished the 40 minutes’ experiment of each 64 participants with the Physiological polygraph (3) Applied ANOVA to identify which Chinese tone causes the most load on humans based on the obtained data in order to improve HCI #2 The Influence of the length and difficulty of the audio on Human Cognitive Load based on Human-Computer Interaction (1) Prepared different audio material and set the length of the context, the difficulty of the context(including Classical Chinese), sex and devices (iPhone or iWatch) as independent variables (2) Designed the flow of experiment utilizing E-prime 2.0 and finished the 30 minutes’ experiment of each 72 participants with the Physiological polygraph Show less
-
-
-
GE Healthcare
-
Medical Equipment Manufacturing
-
200 - 300 Employee
-
OMLP EID(Intern)
-
Jul 2017 - Sep 2017
OMLP= Operation Management Leadership Program NPI= New Product Introduction EID= Early Identification Internship Program Major Tasks: #1 Established Non-CT Organizational Process Asset of 7 programs and 40 configurations and grasped firm knowledge in PRD (Phase Review Discipline) and Vascular, X-ray product #2 Lead solving a new X-ray's IPU (Initial Product Unit) issues, identified, tracked and made improvement in 20 issues in MFG, ENG, MTL, and Packing department #3 Track a new X-ray's product's production plan and ensure MTL readiness for Q3 #4 Support FDR (Formal Design Review) of an upgrade for a Vascular machine, mainly finished IMQA and updated MFG related lists Show less
-
-
-
Chinese Academy of Sciences
-
China
-
Research Services
-
700 & Above Employee
-
Algorithm Engineer Intern
-
Jan 2017 - Mar 2017
#1 Engineered algorithms including Random walk, Pagerank, K-step Markov, SVM and some semi-supervised algorithms to calculate the relevance of diseases, miRNA, drugs, and targets, followed by LOOCV and 5-fold cross-validation #2 Applied Network-based Random Walk with Restart on a heterogeneous network with data form Human Microbe–Disease Association Database (HMDAD) to predict drug-target interactions with the accuracy of 80% #3 Developed a computational model by combining RW with K-means cluster analysis to predict human microRNA-disease association, beating baseline by 60% better Show less
-
-
-
Peking University
-
China
-
Research Services
-
700 & Above Employee
-
Researcher in China Health and Retirement Longitudinal Study(CHARLS)
-
Apr 2015 - Aug 2015
CHARLS is the Project of The National Natural Science Foundation of China Major tasks: #1 Participated and organized a 3 month's training courses, accomplished physical examination and data collection of approximately 300 families based on UNIX about their Health and Retirement problems #2 Coordinated the local departments and residents of 6 villages in Sichuan to move forward, resolved potential conflicts during the research CHARLS is the Project of The National Natural Science Foundation of China Major tasks: #1 Participated and organized a 3 month's training courses, accomplished physical examination and data collection of approximately 300 families based on UNIX about their Health and Retirement problems #2 Coordinated the local departments and residents of 6 villages in Sichuan to move forward, resolved potential conflicts during the research
-
-
Education
-
University of California, Berkeley
Master of Engineering - MEng, Industrial Engineering & Operation Research (Data Science) -
Beihang University
Bachelor's degree, Industrial Engineering -
Beihang University
Second Bachelor Degree, Mathematics and System Science -
King's College London
Strategic management -
Wuhan No.2 Middle School
High school, Science