Shruti Satrawada Email & Phone Number

Santa Clara, California, United States

- Part of the Network Monitoring team at Oracle Cloud Infrastructure (OCI), ensuring seamless performance and reliability of cloud-based networks. - Proactively producing surveillance protocols, resolving anomalies, and optimizing performance for a global customer base.

New York, United States

- Use machine learning as a part of the Growth Notifications AI team at Meta to personalize the notification experience for billions of users- Implementing a model to better predict the best hours to send notifications based on personalized experiences- Using transfer learning to utilize large data sets to build region-specific notification prediction models

Berkeley, California, United States

- Worked as part of the Student Equity Coalition to derive data-driven mathematical insights- Analyzed large datasets to promote equality in higher education through metrics of diversity and inclusivity- Worked to develop and release an open-source toolbox for assessing equity in higher education

Seattle, Washington, United States

- Developed a Dataswarm pipeline in PHP to ingest and analyze data regarding potentially malicious Facebook content on a daily basis.- Employed weak supervision techniques to improve content integrity and prevent malicious curators from hurting the user experience.⁃Created heuristics using text analysis algorithms, content metadata, and other features to derive probabilistic labels for millions of data points.

Santa Clara, California

- Designed 20+ machine learning models, specifically convolutional neural networks (CNNs) in TensorFlow and Caffe- Developed a Python unit test framework to test computer vision deep learning frameworks for autonomous vehicles- Created a user-interactive Jupyter Notebook for Ambarella’s process of porting a mobileNet machine learning model onto a SoC

Berkeley, California

- Trained machine learning models to predict compounds, applications, and relationships between materials using >3 million abstracts - Created a database that spotlights tables and figures that are non-discoverable through search engines by scaping >3 million papers - Designed a comprehensive database for each material and its properties from >3 million research papers

San Francisco Bay Area

http://urap.berkeley.edu/urap-summer/detail/4043Deep Learning for Billion-pixel Digital Pathology Analysis to Improve Diagnosis of Alzheimer's DiseaseIn the Grinberg Lab I worked on the AVID project where the ultimate goal was to alter the PET scan to be able to identify Alzheimer's disease earlier in patients. My team works to scan high-quality images of slices of the brain. We then stitch these images together and use machine learning to identify concentrations of tau, a protein linked to Alzheimer's, in the human brain. One of our final products will be a heat map of the analyzed brains in order to display the concentration of tau. This will be utilized in the analysis of PET scans to diagnose Alzheimer's disease.

Stanford University

- Designed a mechanism for suturing without damage or need for surgical removal. Presented device with an animation of the complete technological process.

Master'S Degree, Computer Science

Bachelor'S Degree, Computer Science

TA for Computational Modeling of the Mind President of UC Berkeley Sahaara Choreographer for UC Berkeley Deewani Ultimate Frisbee Club Team

Education record

Activities and Societies: Vice President of Science Olympiad, Assistant Teacher at Madhuri Kishore Dance School, Competitor in Tennis.