Nithya Sampath Email & Phone Number

Seattle, WA, US

Pittsburgh, Pennsylvania, United States

• Depth Perception for Autonomous Wheelchairs

Redmond, Washington, United States

• Developed new prompts for summarization and theming of ask-me-anything (AMA) events in Viva Engage, a social network serving over 10 million monthly users which connects employees of all levels within a company.
• Created a post-event summary card, recap button, and modal to display AMA summaries to users with React and TypeScript.
• Integrated the summarization API into the backend using Ruby and GraphQL, supporting queries for both single-thread and multi-thread summarization.

• Created a chatbot incorporating several APIs for the beta launch of the Mission Manager, serving 18,000 users by assisting and guiding developers as they engage in coding challenges.
• Planned and constructed an end-to-end system which performs facial recognition, iris detection, and binary classification to detect if a user is cheating during a challenge with 90% accuracy.

Pittsburgh, Pennsylvania, United States

• Planned & led sessions for 21-127 Concepts of Mathematics, a challenging proof-based course.
• Formulated problems based on lecture content to solve during 2 weekly, hour-long sessions.
• Taught topics such as proof structure, induction, etc., serving more than 50 students / semester.

Seattle, Washington, United States

• Researched and defined the car damage detection problem in terms of different CV Problems, including depth estimation, feature extraction, anomaly detection, and finally a combined detection and segmentation problem.
• Extended and customized the Mask R-CNN implementation to train on over 100,000 car images from real customer data, resulting in damage detection with over 80% accuracy.
• Created a small binary classification CNN for instructional overlays, which accept/reject customer-uploaded images based on clarity, illumination, distance, etc.

Pittsburgh, Pennsylvania, United States

• Working in the MultiComp (Multimodal Communication and Machine Learning) Lab to create a new model to learn the latent feature space of a multimodal dataset, specifically one which has the three language, visual, and.
• Creating a pipeline for an artificial dataset with controlled information sharing across modalities.
• Utilizing the latent feature space as a method of generating missing modalities (applications such as generating language corresponding to audio) and predicting data of one modality given data from a different modality.
• Investigating the level of information sharing between modalities for a given multimodal dataset, particularly by determining how accurately a singular modality versus a combination of modalities can reconstruct the.

Pittsburgh, Pennsylvania, United States

• Helping students master math concepts for several hours weekly in various courses, ranging from elementary math to algebra, by working through a variety of lessons and projects together.
• Fostered relationships with students of diverse backgrounds and developed important skills in communication.

Durham, North Carolina, United States

• Held tutorials and problem-solving sessions for AP Physics C and assisted students through difficulties experienced during labs.
• Answered additional physics questions and helped students with physics problems outside of class.

Chapel Hill, North Carolina, United States

• Researched energy availability and consumption patterns in the Embedded Intelligence Lab in order to quantify the reliability of different types of harvested energy, specifically solar and radio frequency energy.
• Constructed circuits, set up sensors, and wrote extensive Python code for experiments.
• Investigated the frequency of energy bursts by generating probability plots of energy availability for solar and radio frequency voltage data.
• Presented at the NCSSM Summer Research Symposium and the SNCURCS conference in November 2019; research paper published in Broad Street Scientific and Journal of Emerging Investigators.

Master Of Science - Ms, Electrical And Computer Engineering

Concentration in AI/ML Systems Relevant Courses: Algorithms for Large-Scale Distributed ML, Deep Learning & Pattern Recognition for.

Bachelor Of Science - Bs, Electrical And Computer Engineering, 3.8/4.0

Activities and Societies: CMU Bhangra, Women in ECE, MultiComp Lab ResearchMinor in Computer Science Relevant Courses: Computer Vision.

High School Diploma

Activities and Societies: Varsity Tennis Team, Bhangra Dance Team, Raas Dance Team, Math ClubHonors/Awards: MathCON Nationals Gold Medal.