Hanmi Zou

Experience

• 

  William & Mary

  Jan 2018 - Aug 2018

  Undergraduate Research Assistant of Optimizing the Matching Between Students and Advisors

  • Managed $3000 scholarship to build a mathematical model that can efficiently pair freshmen with mentors based on the Transportation Problem Models, facilitated collaboration between mentors and students using a consultative approach• Crafted and executed operational strategies to uncover insights about the model to executive audience, Applied AMPL and NetworkX package in Python to build mathematical model to maximize the satisfaction degree between freshmen and mentors• Achieved the outcome that 80% of the students got a mentor that shared their first ranked major; almost 100% of them got mentors that shared common majors; 72% of them even got two or more majors in common with mentors• Authorized the software prototype to the administration department, advised on best practices of software thus enabling them to make sound business decisions, chosen as the 2019 Joint Mathematics Meeting Student Poster Session Award Winner • Prepared to publish the research on Involve a Journal of Mathematics, a research journal dedicated to mathematical research by students Show less

• 

  IGEM Competition

  Mar 2018 - Oct 2018

  Modeling & Visualization Engineer, 2018 The International Genetically Engineered Machine Competition

  • Designed an information decoder circuit using an incoherent feed-forward loop (IFFL) to convert frequency-encoded information into amplitude-encoded information, thus enabling synthetic biology to interpret dynamics signal in natural system • Implemented a mathematical model to incorporate ordinary differential equation (ODE) and stochastic factor using ODE package in python, explored parameter spaces and performed model fitting with Markov Chain Monte Carlo (MCMC) method• Visualized the superiority of IFFL decoder circuit over the naïve circuit by heat maps, created and maintained the team webpage, presented the results at the iGEM Mid Atlantic Meet Up hosted by the University of Maryland• Won the Gold Medal at Giant Jamboree research competition in Boston and interviewed by Daily Press Show less

• 

  Huazhong University of Science and Technology

  Jun 2019 - Aug 2019

  Data Scientist(Internship), School of AI & Automation, Huazhong University of Science and Technology

  • Designed a multi-label heart-diseases-classification system algorithm using multilayers convolutional neural networks (CNN) to predict heart diseases symptoms with variational auto-encoder (VAE) and generative adversarial network (GAN) • Assigned labels to different diseases, performed training on regularized dataset of ECG signals, built CNN with Adam optimizer and achieved 97% accuracy of the classification results• Denoised the dataset by applied Auto-encoder and linear digital filters to improve the accuracy of detecting QRS-complex to predict next onset of heart disease, extracted different features and observed possible trends through denoised dataset • Compared the proposed methods of GAN and expectation maximization algorithm (EM algorithm) to the baseline methods, i.e., clinicians’ diagnosis, to gain better insights about prediction Show less

• 

  Zebra Medical Vision Ltd

  Sept 2019 - May 2020

  Machine Learning Research Intern (Part-time) working on Capstone Project with Zebra Medical Vision

  • Develop a large-scale machine learning triaging algorithm to visualize Intra-Abdominal infections in CT Scans, an important finding that provide an early alert to radiologist to prioritize patients with acute condition, achieve 91% precision score• Effectively designed NLP weak labeling technique to distinguish acute conditions and non-acute cases, achieved 98.08% precision, 93.62% recall, 99.18% specificity and 97.85% accuracy on manually labeled validation set • Refined NLP by analyzing challenging false positive/negative case, preprocessed 1.6 million CT scans input by choosing desired organ region and segmentation, architected convolutional neural networks (CNN) visual classifier with high precision• Partnered with product managers and clinicians to transform algorithm into business insights, prototyped and productionized algorithm features to interpretable Proof of Concept (POC), provided effective traceability metrics to executive audiences Show less

• 

  Berkeley Lab

  Feb 2020 - Jul 2020

  Deep Learning Research Associate

  • Collaborated with Department of Energy (DOE) and Department of Veterans Affairs, combining high-performance computing and machine learning techniques to improve mental health for veterans by extracting patterns from millions of health records • Explored positive correlation between suicide and altitude by examining 30 years of county-specific mortality records obtained from Centers for Disease Control and Prevention(CDC), resulting in statistically significant difference with p < 0.0001, this novel risk factor could be essential for suicide prevention• Conducted geospatial data exploration and visualization on county-level suicide information from U.S. Geological Survey database using SQL and Python (i.e. Pandas, Matplotlib) to further analyze the suicide rate on each county• Participated in organizing and cleaning database for mortality data associated with different factors such as altitude, education level and medical conditions ranged from years 1979 – 2019 for future public usage Show less

• 

  Eli Lilly and Company

  Sept 2020 - Mar 2021

  Statistical Programmer

  Assigned by Brightech International• Provided technical support for seven key pharmaceutical projects, advancing products to phase 3 and FDA submission.• Developed statistical analysis plans and SAS programs for data analysis, identifying market trends, demographic insights, and ensuring CRF (Case Report Form) compliance with protocol guidelines.• Collaborated closely with Biostatistics and Data Management teams for clinical data verification, CRF annotations review, and quality control through active communication with programming and statistics leads. Show less

• 

  Johnson & Johnson

  Jun 2021 - now

  • Led integration of 112,826 HCP (Health Care Professional) master data from 20+ Omni-Channels, enhancing J&J data governance and maturity.• Implemented a patient follow-up management solution, achieving 100% compliance for Mentor Product license renewal and integrating with J&J data platforms for R&D insights.• Managed the J&J Wound Closure WeChat HCP portal, growing to 17,000+ monthly active users and introducing 19,500+ new users, with total video playtime reaching 9,720 hours in 2023.• Delivered a digital customer experience product catalogue and data dashboard, significantly improving HCP engagement and operational efficiencies. Show less • Joined as ITLDP; led Janssen R&D's (Pharmaceutical sector in J&J) first rotation, developing Janssen Easy Development Intelligence system to monitor competitors and gather insights, supporting informed business decisions.• In second rotation at MedTech China, accelerated digital transformation for Biosense Webster (BW) as project manager; implemented Carto Journey model, supporting EP physician growth and delivered BW data platform, offering flexible data services.• Played key role in J&J ITLDP Campaigns; led teams in 2022 J&J Technology SharkTank and Technology Summit, bridging business and technical teams, and providing technology thought leadership. Show less

  • Senior IT Data Analyst

    Jul 2023 - now

  • ITLDP Management Trainee (Information Technology Leadership Development Program)

    Jun 2021 - Jul 2023