Victor Tran Email and Phone Number

Using predictive modeling, statistics, Machine/Deep Learning, and other data analysis techniques to explore, and extract insights from structure and unstructured data.Planning, designing and developing tools and models for data analysis.Developing software, algorithms and applications to apply mathematics to data, perform large scale experimentation and build data driven apps to translate data into intelligence, utilizing efficient data loading and data augmentation techniques.Designing and development of data products using selected tools and frameworks as required and requested.Performing tasks such as writing scripts, develop pipelines, SQL queries, analytics reports, etc.

As a member of the technology manufacturing group I support operations and technology development. I own a portion of the purification process equipment and am responsible for keeping them running well within specifications. Scope includes but is not limited to: driving process improvement and cost-savings projects, commissioning and starting up new equipment, troubleshooting process alarms, writing detailed procedures for non-standard work, engaging is task forces, interfacing with vendors and contractors as well as collaborating with interdisciplinary groups and counterparts across the virtual factory to address the challenges of supporting an ever-evolving industry and purification process.

Researched, developed, and tested anomaly detection algorithms for application in full scale manufacturing settings using semi-supervised and unsupervised machine learning models. Algorithms used for performing automated weld defect detection, enabling QA/QC at a level that is normally economically infeasible. Worked with large data sets of ~1,000,000 rows and ~150 features using common mathematical and visualization tools such as: Numpy, SciPy, Pandas, Matplotlib, and Seaborn.Pre-processed a large volume of structured and unstructured data, independently learned welding domain knowledge necessary to perform educated feature selection/engineering, and used a combination of clustering techniques (K-means++, DBSCAN) and dimensionality reduction techniques (t-SNE and PCA) for EDA of high-dimensional data and structure discovery.Used PyTorch and Scikit-learn frameworks to train multiple state-of-the-art anomaly detection models using including auto-encoders, isolation forest, one-class SVM, evaluating classification performance using AUROC, F1, and confusion matrices and performed model validation and benchmarking on public datasets.

I worked as a teacher’s assistant in UC San Diego’s chemical engineering department during my MS program. I communicated and coordinated with the professor and the other TAs to evaluate students’ homework, exams, and projects. As a TA one of my goals was to ensure that grading was done in a consistent manner and in a way that accurately assess the students’ learning outcomes. I also did my best to provide valuable feedback tailored to each students’ unique needs which is arguably the most important role of a TA. Senior students were assigned a design project spanning two quarters which involved regular submission of technical reports. The role of TAs were especially critical during this project as we acted as the primary source of feedback and were expected to see the assigned project groups through from beginning to end, requiring us to follow the developments closely and demonstrate mastery of the material.TLDR: Graded assignments for upper-division chemical engineering courses

I was a one-on-one algebra tutor at Foothill College’s for two years, tutoring 1 or 2 students each quarter. Each session would begin with a discussion to get me caught up on materials covered in class since the last meeting and decide how to make the best of use of our time. The bulk of the sessions were spent identifying and filling gaps in the tutee’s understanding, preparing for upcoming exams, and sharing strategies for academic success. At the end of each session we would summarize what was accomplished and agreed on a plan for the next meeting. TLDR: Meet regularly with tutees to answer technical questions, provide preparation & guidance for exams, review & practice material.