Sid Dhawan Email & Phone Number

San Jose, Ca, Us

San Francisco, California, Us

San Francisco, Ca, Us

Identified methods to efficiently use Machine Learning to automate the creation of doctor’s notes. * Based on customer feedback, created an algorithm to automatically detect and correct mistakes in ASR transcriptions, thus improving accuracy of T5 summary generation. This improved usability by allowing doctors to dictate their notes.* Updated T5 summary generation mechanism to create variations in opening words and sentence lengths. This improved readability within generated doctor’s notes. * Implemented data version control; built an automated pipeline to transform raw data to training inputs for our NLP models-incorporated parallelism, integration with databases, and sensitive handling of PHI.

Santa Clara, California, Us

Worked to improve and monetize the Hyperseg algorithm, which performs image segmentation with State-of-the-Art precision (≥ 90 mean IoU). * Built a GUI to allow Hyperseg to be easily used by non-technical animators. Managed packaging and distribution of software to customers using Docker. This saved many hours of manual labor for the client.* Utilized “Backpropagation Refinement Scheme”, (Dang, Kim,CVPR ’19) to improve Hyperseg’s segmentation quality via inference-time refinements. * This work was accepted for publication at CVPR ’21. * Gained experience in PyTorch and Tensorflow.

Cambridge, Massachusetts, Us

Enhanced and maintained big data infrastructure, allowing analysts to generate actionable insights from hospital EHRs. * Rewrote the data lake ingestion pipeline using Spark and AWS. As a result, hospital data could be incorporated into the data lake at up to 10x the previous speed and with far fewer crashes, thus saving hospital and company time.* Collaborated cross functionally with teams and clients to gather data (for ~15 EPIC-Clarity Hospitals) and integrate into CLQ’s clinical measures system. This allowed the company to acquire expertise in providing services to EPIC-Clarity clients, and nearly doubled the user base.* Utilized AWS Athena and SQL to isolate data for patients with specific cancers. This initiative generated revenue and provided tens of thousands of patient outcome data points to improve drug discovery research.

San Jose, California, Us

Built a large-scale Machine Learning algorithm (Belief Propagation) in python's “Graphlab” library to detect potentially malicious IP addresses. This is a useful addition to Symantec's internal and commercial network security algorithms.

New York, New York, Us

Designed and implemented a Statistical Quality Control algorithm in R, based on Bayesian Statistics, to analyze 24-dimensional flavor data from a collection of users to identify significant variations artisan food products. This was a crucial step in allowing AFS’s quality control system to be ready for use by potential clients

Master'S Degree, Computer Science

Bachelor'S Degree, Computer Science