Harshvardhan Gupta

• Designed and implemented a robust ETL pipeline using Python and SQLAlchemy that processed over 18 million claims annually, streamlining data integration and improving overall system efficiency.
• Developed an innovative statistically assisted trigger system using Flask and PostgreSQL to identify potential savings opportunities on bills, resulting in an average savings of $350 per person for clients.
• Actively participated in agile software development processes to produce timely and high-quality deliverables for partners, fostering strong collaboration and communication across teams.
• Collaborated with front-end developers to seamlessly integrate back-end services with front-end interfaces, ensuring a consistent and efficient user experience.
• Engineered a resilient infrastructure using Terraform and AWS, capable of handling extremely elastic loads during open enrollment periods, maintaining system stability and performance.
• Designed and built efficient RESTful APIs using Flask and Python to support more than 1,000 concurrent users, optimizing system performance and scalability to meet growing demand.
• Proficiently profiled and optimized critical SQL queries in PostgreSQL, achieving performance gains of up to 100x, and significantly reducing database latency and resource consumption.

• Conducted an in-depth research project on the feasibility of using Recurrent Neural Networks (RNNs) for Quadratic Residue Detection, working independently under the guidance of a supervising professor.
• Executed compute-intensive experiments on a weekly basis, systematically evaluating model performance and gathering critical insights to inform future iterations.
• Analyzed the pros and cons of various parameters, continuously refining and enhancing machine learning models to optimize performance and achieve desired outcomes.

• Selected as a member of the prestigious BA2 Cohort program with an acceptance rate of less than 5%, demonstrating exceptional skills and expertise in advanced analytics and machine learning.
• Developed a distributed Speech Analytics software for vernacular languages using transformer networks, with early prototypes adopted by several clients and achieving state-of-the-art performance (NLP, Python, Pytorch, Spacy, Flask, TensorFlow, AWS).
• Designed and implemented High Fidelity Steganography for streaming content using novel Resnet-style CNNs, tailored to work with JPEG compression through the use of JPEG-aware models (Python, Flask, Pytorch, OpenCV).
• Created a unified framework to efficiently distribute machine learning models as REST APIs, streamlining model deployment and integration (Python, Flask).
• Collaborated closely with stakeholders to design and develop efficient development cycles, ensuring timely delivery of high-quality solutions and fostering strong working relationships with clients.

• Developed a modern Flask REST API framework to enable efficient development in a performance-intensive workflow, supporting up to 10k real-time users and ensuring seamless system scalability.
• Designed and implemented custom secure APIs to interact with over 50 companies using novel asymmetric encryption systems and signed payloads, enhancing data security and integrity.
• Built a user-friendly tool for partners to access key metrics in real time, fostering transparency and effective collaboration with external stakeholders.
• Collaborated closely with external engineering teams to ensure successful integration, facilitating cross-functional communication and aligning project goals

(Python, Flask, Django, Machine Learning, RESTful APIs).

• Successfully delivered large-scale machine learning models as Flask APIs for various freelance projects, demonstrating expertise in efficient model deployment and integration.
• Applied novel techniques to mitigate the risk of image leakage, ensuring data security and privacy across diverse projects.
• Collaborated closely with stakeholders to monitor progress and ensure timely completion of project milestones, fostering strong working relationships and maintaining a high level of customer satisfaction.

• Developed an innovative Android app that stitches together questions from a question bank to generate realistic question papers with solutions, enabling students to effectively practice for exams.
• Integrated supporting tools such as calculators, dictionaries, and cheat sheets within the app to provide a seamless and comprehensive learning experience for users.
• Designed and built an internal tool for efficient and secure data entry, streamlining the process of inputting information into the question bank and ensuring data accuracy.

Main skills used: Android Development, Java, AWS