Advisor: Leanne Hirshfield• Designing and implementing Transformer based 4D fMRI architecture for task-based fMRI modelling. (ongoing)• Implemented the the measurement of correlation between deep brain fMRI and fNIRS signals using Pearson’s correlation, Dynamic Time warping and MdRQA with its extension to model FMRI signals using FNIRS by encoder decoder architecture.(paper in progress)• Contributed to the proposal of Neuro-Inspired deep reinforcement Learning based Human-Agent Teaming architecture withreward shaping using Neural signals.• Developed transcription system with diarization using Whisper AI and pyannote for experiment/study recordings (iSAT)outperforming Google Speech.

Advisor: Lei Yuan• Modeled child’s learning using Image Captioning encoder-decoder model (using CNN and LSTM) for English and Chinese language to showcase the learning of generative principles by the model and its comparison with child learning.

Student Assistant for Course CSCI 3308 - Software Development Methods and Tools

• Designed and implemented workflow (orchestrator) service for Policy Simulation based on change set with staged invocations, notifications and change approval using authorization to determine the impact of complex policy changes across WorkEventssystem. Led multiple teams (3) for requirement gathering, product feature finalization, design and implementation. [Python,Cloudformation, Lambda, Step Function, API Gateway, DDB]• Designed High Level Design for Policy Simulation Service covering key design choices of simulation environment,parallel workflow invocations, locking features for maintaining consistency and identifier injection for tracing invocation of all the downstream services. Also, Designed Low Level Design for the service covering orchestration using Step Function with API Gateway, Wait for Task Token Async design, Auto DB Sync, DynamoDB Table design and API Design. Presented complete design and got it approved for further implementation.• Implemented the packages for Policy Simulation Service with base code, required abstractions and modularity. Developed infrastructure for components using Cloudformation (code as infra) to enhance extensibility and reuse along with automated service monitoring, alarms and CI/CD deployment. Created generic exception handling package using decorators which was used across projects.• Mentored 3 SDEs about organizational services, tools and software design and development practices at Amazon.• Supported as TA for Machine Learning Bootcamp and Deep Learning Bootcamp at Amazon.

• Designed and implemented VNHO payment automation portal (along with complete test suite) and launched across multiplecountries. Implemented authentication and authorization for the API services and UI. [Java, Ruby, React, Redux, Enzyme, AWSDynamoDB]• Designed and implemented the APIs (with required unit and integration tests) for WEReporting Service using restfulconventions. Developed the portal for report management and download. Identified and implemented service monitoring alarms for the WEReporting Service and created wiki dashboard (codified) for them. Service was part of WorkEvents product and was launched successfully across multiple countries. [Java, Python, React, Redux, Ruby, Wiki, Monitoring Services]• Integrated Population Service APIs for fetching employee details in Pay Portal. Implemented Step Function Activity worker to run the service in MAWS through NAWS Step Function. [Java, AWS Step function]• Identified the solutions for data streaming of WorkEvents data and analyzed their capabilities (including Amazon Kinesis as well as change tracking of the SQL DB). [Python, Amazon Kinesis, Amazon RDS]• Created quick starter code to deploy dynamic websites using NodeJS, React, Redux served through Cloudfront which was used across the organization to build further products faster with the reuse of the base services. Conducted Tutorial sessions on "React and Redux Design Principles and Use" to help build websites faster across the org.• Mentored 2 SDE interns on common AWS Services as well as organization owned services in various tech stacks to help them to build new services.

InternshipDevelopment of Android application.