My Work in Action
Student Research Assistant
Rehabilitation Engineering Design Laboratory, OSU
Sep 2024 – PresentSpearheading the development of a real-time diagnostic system for wheelchair breakdown detection, I’ve defined project milestones and am currently executing diagnostic algorithms using cloud-based tools and APIs to enhance system accuracy by 25%. By integrating sensor data with cloud technologies, the system aims to enable seamless analysis and faster detection of breakdowns, with a projected 20% improvement in response time. Additionally, I’m developing an Android application that integrates Google APIs (Distance Matrix, Cloud Storage) to automate sensor data collection, streamline the process, and provide real-time insights into wheelchair performance. This integration is expected to significantly improve operational efficiency, reducing delays in breakdown detection and enhancing the overall user experience. Through this role, I’m strengthening my skills in real-time data processing, cloud integration, and mobile app development, while also gaining valuable experience in problem-solving and project management.
Undergraduate Teaching Assistant
Department of Computer Science, OSU
Sep 2024 – Dec 2024Graded 400+ assignments in Object-Oriented Programming and C++, providing personalized feedback to help students grasp key concepts and improve their programming skills. I identified common challenges and addressed individual problems, offering targeted explanations that significantly boosted student understanding and performance. This experience enhanced my problem-solving and teaching skills, allowing me to break down complex topics into simpler, digestible formats. Working with diverse students also sharpened my communication abilities, enabling me to tailor my approach to individual learning needs.
Research Assistant
Jan 2023 – Jul 2024Engineered innovative data collection systems to support cutting-edge research in mental health. For the suicidal ideation project, I developed a robust pipeline to gather and preprocess data from over 10,000 Facebook posts, utilizing text, emoji usage, and user interaction data to identify patterns linked to depression and suicidal thoughts. This effort directly contributed to the creation of a predictive model aimed at suicide prevention in South Asia, addressing regional challenges in behavioral patterns, socioeconomic factors, and language nuances. In parallel, I spearheaded the data analysis for the infodemic news analysis project, where I collected and processed a 4-year dataset of news articles from India, Bangladesh, and the UK. Using NLP techniques, I identified trends in misinformation during public health crises, providing actionable insights to enhance global public health response strategies. Additionally, I collaborated on building a machine learning model for stress level assessments based on psychological metrics (PSS, BFI-2), improving the accuracy of mental health diagnostics. This work reinforced my technical skills in Python, NLP, and data analytics, while also deepening my understanding of how data-driven approaches can address critical social challenges.
Data Scientist
Feb 2023 – Sep 2023Orchestrated the development of a robust sales forecasting model, achieving 92% accuracy using a customized FB-Prophet approach. This breakthrough significantly enhanced client decision-making and operational planning, directly contributing to improved forecasting precision, reduced operational costs, and increased client satisfaction. Collaborating with 10 stakeholders, I translated diverse business requirements into a scalable SaaS-based Point of Sale (POS) system, leveraging Django and React to seamlessly integrate real-time sales and returns reporting via REST APIs, which streamlined internal operations and reduced reporting time by 30%.
To further enhance client communication, I developed and deployed a WhatsApp bot service, providing clients with real-time updates and improving engagement. By utilizing SQL, I optimized data extraction and preprocessing workflows, which not only improved model accuracy but also led to faster and more efficient data handling. Leading the team to deliver the POS system, I mentored 20 new employees in technical skills like Python, Django, and PostgreSQL. This not only fostered a culture of continuous learning but also strengthened the team’s overall technical proficiency.
This experience deepened my ability to bridge technical expertise with business outcomes, strengthening my problem-solving, leadership, and mentoring capabilities.
Software Developer Intern
Jul 2022 - Dec 2022Led the development of a high-precision Named Entity Recognition (NER) system using Regex and Stanza, enabling the extraction of location and date-time entities from vast multilingual datasets. This improved lead identification, increasing relevance by 25% and enhancing data processing efficiency for a leading geospatial mapping company. I also engineered a multi-label text classification model, significantly improving the categorization of leads by refining the algorithm's precision, resulting in faster, more accurate categorization processes. Additionally, as part of a Sofathon, I designed and implemented an advanced image translation system using pix2pix and GANs, accelerating image processing times and reducing overhead by 40%. My approach leveraged TensorFlow and PyTorch, optimizing the models for real-time deployment. These innovations not only streamlined data updates but also enhanced the overall user experience, making it more responsive and scalable. The 8/10 score for strategy and implementation in the Sofathon further validated the effectiveness of the solutions. Through this work, I gained expertise in NLP, deep learning, and AI model deployment, equipping me to deliver impactful, scalable solutions that drive efficiency and performance across dynamic environments.
Software Engineer Intern
May 2022 - Jun 2022Optimized image processing pipelines by leveraging Python and MLflow, increasing efficiency by 25%. I implemented a robust solution by automating data preprocessing, feature extraction, and model training. By conducting A/B testing, I evaluated various model architectures, selecting the most effective approach based on performance metrics like precision and processing time. This optimization led to faster data analysis and more consistent model predictions, enhancing the speed of client deliverables. Through this process, I gained hands-on experience in model selection, fine-tuning, and real-time deployment, learning how to balance computational efficiency with prediction accuracy for scalable solutions.
Al&ML Intern
Integrated Active Monitoring Pvt. Ltd., India
Oct 2021 - Apr 2022Spearheaded the automation of critical surveillance processes, cutting manual efforts by 80% and enabling real-time truck detection and tracking using PP-Yolo and Byte-Tracker models. Enhanced security operations by automating CCTV monitoring with OpenCV, implementing features to detect video blur, scene changes, time mismatches, video loss, and hard disk health issues, improving system reliability by 40%. Developed an OCR system with PaddleOCR, boosting skewed text recognition accuracy by 25%, and built predictive models for age, gender, and emotion detection, increasing image analysis accuracy by 30%.
Designed and deployed an end-to-end data processing pipeline integrating a ReactJS UI, FastAPI backend, MongoDB for data retrieval, RabbitMQ for efficient messaging, and Docker for containerization, accelerating data processing speeds by 50%. These innovations significantly improved operational efficiency, ensuring faster alerts and greater reliability in real-time tracking and security monitoring. This project reinforced my expertise in computer vision and automation while delivering measurable business impact and showcasing my ability to create scalable, high-performing solutions.
Summer Intern
The Robotics Forum (TRF) - The official Robotics club of VIT Pune
Jul 2020 - Sep 2020Designed and controlled a robot for object picking, dropping, and lane-following, applying my knowledge of robotics, sensor integration, and real-time system control to create a functional prototype. My responsibilities included programming the robot to navigate autonomously using sensors for object detection and path tracking. I successfully integrated sensors with the control system, allowing the robot to perform tasks with accuracy and efficiency. This experience provided me with valuable hands-on knowledge in embedded systems, sensor calibration, and real-time decision-making algorithms. The project enhanced my problem-solving skills and ability to work under constraints, fostering a deep understanding of robotics systems and their real-world applications.
Click or tap a role to read more details.