NewHero

Accelerate the transition from human-as-tools to human-as-creators

CEO and Co-Founder

Artflow

One-stop AI content creation platform for images, videos and stories.

Co-Founder and CEO

Wayfair

- Led the Image Synthesis project that unlocked training possibilities for multiple downstream applications: shape/material/color-aware visual search,  camera perspective estimation, material segmentation
- Backfilled 200k 3D models with shape/material metadata and indexed their similarities, reduced scene population time by 24% by allowing stylists to find what they are looking for. Increase 3D model reuse rate by 17% by surfacing 3D assets that was previously not tagged. Also created novel storefront opportunities to allow customers search for products intuitively using color, shape and material
- Led a cross-functional team on 3D material segmentation and retrieval, for the first time, allowing artists and modelers to identify duplicated materials and reuse approved materials
- Unlocked 70% of the color options on Storefront 3D models that resulted in 6x more 3D models created. Increased customer engagement (selection + movement) in storefront 3D experience by 162% 
- Judge and mentor at MIT Media Lab Hacking Arts for the 2018 and 2019 events.

Tech Lead @ Data Science - Vision and Speech Platform - 3D Innovation

- Led the Scene Graph project, for the first time, allowing internal stakeholders to conduct 3D asset management in a visual and interactive knowledge graph manner
- Led cross-functional team on gallery wall synthesis project, automated wall art recommendation and placement, deployed on Magic Leap One
- Synthesized hundreds of thousands of synthetic furniture images using 3Ds Max, improved the Visual Search algorithm on low performing classes by a large margin
- Built camera perspective estimation model with realistically synthesized training images, which replaced manual perspective estimation, led to $700k cost savings/year

Senior Machine Learning Engineer @ Data Science - Vision and Speech Platform

- Built functional prototype for single image-based 3D topology/texture generation, presented to company co-founder
- Reduced photogrammetry masking time by 300% via neural network-based automatic object masking
- Led multiple research collaborations with MIT, Umass Amherst and UMass Boston, resulted in multiple academic papers and working prototypes, presented the results to leadership.
- Mentored 1 PhD intern on the topic of AI-assisted Interior Design

Research Engineer - Wayfair Next

Boston Computational Creativity Meetup

Local meetup with 400+ members focusing on sharing latest progress in the intersection of AI and Creativity.

Founder

DePaul University

• Created a dashboard for visualizing students’ academic performance across 14 university departments, the goal is to provide a holistic view for divisional leadership and staff members to better prioritize resources and set focal point, and ultimately tie in academic performance with student retention, and enables effective intervention.
• Improved student retention and on-time graduation via predictive modeling
• Identified gaps in college persistence and degree completion by mining student data
• Improved data validity for 56 student groups by planning and carrying out data update schedules
• Provided trainings on technology platforms for non-technical staff members

Graduate Assistant for Technology and Information Management

Analyzed Chicago Public Libraries' book circulation, neighborhood demographics, promotional events, social media mentioning as well as book text data, built predictive models, and developed multiple interactive applications to support "One Book One Chicago" program to make data-driven decisions on future book choices and event planning. 

Our project page:  http://dh.depaul.press/reading-chicago/

Research Assistant

VisMed3D - Biomedical Design, Printing and Consulting

Collaborated with three other colleagues on building an information extraction web application, including automated news article searching and scraping, article relevancy ranking, named entity recognition and highlighting, database design and server-side deployment.

Data Engineering Intern

The University of Iowa

•  Cleaned, aligned and normalized biometrics data such as heart rate, pupil dilation, eye movement and facial expression records into an analysis-ready form
•  Delivered quantitative and qualitative analysis of 130+ experiment subjects that discovered underlying correlations between certain physiological status and emotional states
•  Developed an algorithm that consolidates collective biometric responses into comprehensive insights about the commercial
•  Extracted 300+ features that used in predictive model building, was able to predict individual-level arousal and valence state 11% more accurate than baseline
•  Diagnosed imperfect data and generated a report that helped avoid bad data collection in the future
•  Conducted ad-hoc statistical analysis and dimensionality reduction on high-dimensional temporal data

Research Analyst

• Worked as a full-stack developer on a NSF-supported project researching on human decision making
• Developed a plug-and-play web app that allows remote participation without Internet access
• Automated the ETL process by enabling loading self-defined research configuration data
• Built a highly customized research configuration page, featured with intelligent form population
• Designed a relational database to house all research‑related data with role‑based security
• Integrated research framework with gamified virtual environment that improved data validity and user friendliness
• Worked with project supervisor to determine the games and report presentation that users might be interested in. Then created novel methods for visualizing and presenting the data in a powerful, fun, and elegant way

Application Programmer/Analyst

•  Engineered 40+ features, intended to characterize and predict high-frequency trading activities
•  Automated a feature extraction pipeline that can be easily applied to other trading days
•  Selected the most relevant features using Genetic Algorithm
•  Mined 60 million stock trading and quoting records via K-Nearest Neighbors, raised prediction AUC to 0.733

Tim's Studio

• Provided financially viable photography services for 300+ college students, 8 small businesses and organizations
• Led a 5-­person team covering all aspects of photography business operation
• Undertook special photo projects in collaboration with student organizations to raise fund and awareness for cultural diversity, environmental sustainability and philanthropic endeavor

HNI Corporation

• Developed and presented 20+ metrics to CIO and IT leadership to evaluate business alignment and IT strategic planning for the Enterprise Architecture Group
• Collaborated with operational teams to identify appropriate data source for metrics
• Managed storage optimization project by leading cross-team meetings, finalizing project charter, prioritizing and implementing project tasks 
• Liaised between cross-functional teams to deliver weekly progress report
• Facilitated IT demand intake process restructuring by liaising project executives and assigning action items
• Performed administrative assistance in finalizing demand types and visualizing intake processes to improve business standardization and agility

Enterprise Architect Intern

User Recommendation in Reciprocal and Bipartite Social Networks--A Case Study of Online Dating

• Preprocessed data through data separation, filtering and matching
• Constructed a small-size social network using NetworkX and Matplotlib to visualize nodes, edges and specific user identities
• Conducted cluster analysis by computing clustering coefficient and finding cliques
• Test and evaluate of social network recommender system

Research Programmer

Building the best AI tools to help personal brand builders thrive

Taplio

Tim Zhang's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Tim Zhang's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Tim Zhang's Linkedin Analytics