Feb 2025 – Mar 2025 · 1 mo · Bengaluru, Karnataka, India

• Led development of a Machine Learning algorithm using Random Forest classifier to predict optimal QBO SKUs based on QBDT customer usage data, achieving an impressive 83% accuracy rate.
• Led development of React-based UIs for the Post-Migration Hub, ensuring a seamless transition for over 400K customers from a Quickbooks Desktop to QuickBooks Online application.

Jul 2023 – Jan 2025 · 1 yr 6 mos · Bengaluru, Karnataka, India

• Implemented REST APIs using Spring Boot with various authorization signatures, enhancing security.
• Implemented a stream processor in Java using Apache Beam to listen to Kafka events, extract parameters, and create setup checklists for users migrating from the QBDT to QBO application, enhancing the migration process and user experience.
• Utilized GenAI to transform common English sentences into Splunk queries, making complex data searches more accessible to non-technical users such as Product Managers and Experience Designers, thereby enhancing team efficiency.
• Integrated a Qualtrics dynamic survey into the system to periodically gather user feedback; channeled feedback to a Slack channel for real-time monitoring and response; leveraged GenAI to summarize daily user survey responses, streamlining insights generation and decision-making processes. Increased the take rate from 10% to 27%.

May 2022 – Jul 2022 · 2 mos · Bengaluru, Karnataka, India · On-site

• Developed dynamic data visualizations in React JS, transforming table data into interactive bar charts and line charts.
• Implemented and tested REST APIs using Java and Spring Boot, ensuring robust and efficient data handling for front end.
• Implemented a chart editor component in React JS to control metrics, dimensions, and visualization types of charts; wrote unit tests using Jest Framework to ensure component reliability and functionality.

Jun 2022 – May 2023 · 11 mos · Kharagpur, West Bengal, India

• Developed mmDrive, a novel system for real-time monitoring of dangerous driving behaviors using millimeter-wave (mmWave) radar sensing, achieving a high accuracy rate of 97% for detecting nine distinct dangerous driving actions.
• Contributed to model architecture and data analysis, including designing and implementing a Fused-CNN classifier that combines range-doppler and noise profiles to distinguish between normal and dangerous driving behaviors.
• Led field experiments with extensive data collection from real-world driving environments, utilizing IMU sensors and radar data to filter road-induced noise, enhancing system robustness against road bumps and other disturbances.
• Published research findings that showcased mmDrive’s accuracy and advantages over traditional vision and wearable-based systems, establishing its potential as a privacy-preserving, on-device solution for driving safety.

Sep 2021 – Apr 2022 · 7 mos · Fremont, California, United States · Remote

• Implemented Cloudflare Workers server code to send conversion events using Facebook Graph APIs, optimizing event tracking and enhancing marketing campaign performance through reliable and efficient serverless functions.
• Developed a custom Airbyte Python HTTP API Source Connector for Xola APIs, enabling automated, scalable, and reliable data integration into data pipelines, enhancing data consistency, and streamlining analytical workflows.

Jun 2021 – Apr 2022 · 10 mos · Kharagpur, West Bengal, India

• Developed ExpresSense, an innovative acoustic-based system for detecting facial expressions on standalone smartphones, achieving a classification accuracy of 75% across basic expressions without requiring camera-based input, thus enhancing user privacy.
• Engineered the core system architecture, using near-ultrasound acoustic chirps (16-19kHz) to detect micro and macro facial expressions, eliminating reliance on external hardware and enabling a cost-effective and privacy-focused solution for real-time expression detection.
• Led data collection and experimentation with diverse smartphone models, conducting real-world testing with multiple participants to validate ExpresSense’s ability to recognize expressions accurately in unconstrained environments.
• Published findings that demonstrated the feasibility of standalone smartphone-based expression recognition, paving the way for potential applications in privacy-sensitive areas like real-time feedback in online education, emotion-based HCI, and assistive tools for individuals with physical disabilities.

Apr 2021 – Jun 2021 · 2 mos · Bangalore Urban, Karnataka, India · Remote

• Developed RESTful APIs using DRF to facilitate seamless communication between frontend and backend systems
• Enhanced API security by integrating custom permissions, ensuring data integrity and protection against unauthorized access.
• Achieved a remarkable 15X performance boost for APIs through strategic optimizations including bulk create or update operations and code enhancements, significantly enhancing system efficiency and user experience