Archit Mandpe

Conducted in-depth data analysis and visualization to derive actionable insights from user behavior and optimize resource utilization.

• Conducted comprehensive data analysis on extensive user data volumes to gauge user behavior, leveraging advanced analytical techniques for strategic insights.
• Addressed and resolved data redundancy issues within user volume data, performing robust feature engineering using PySpark to optimize data quality and analytical efficiency.
• Generated critical insights and visualized complex user resource usage patterns using Tableau Software, informing data-driven decision-making.
• Developed and implemented a robust forecasting model to accurately predict when user data would reach threshold storage limits, enabling proactive resource management.

Designed and implemented a comprehensive machine learning system for user expenditure forecasting, enhancing data integrity and model performance.

• Designed and implemented a comprehensive end-to-end machine learning system to accurately forecast user expenditure based on historical transaction data.
• Leveraged PostgreSQL for robust storage of user transaction data, performing extensive exploratory data analysis and transformations to select key features, significantly improving model accuracy and ensuring data integrity.
• Streamlined machine learning model development, testing, and inference pipelines by effectively utilizing MLflow and Weights & Biases.
• Architected a high-performance and lightweight FastAPI server with RESTful APIs, enabling efficient and scalable model inferencing.

Developed and deployed a real-time audio transcription system leveraging state-of-the-art ASR models and optimized for performance.

• Designed and implemented a real-time audio transcription system utilizing state-of-the-art ASR models from Hugging Face pipelines.
• Integrated Wav2Vec 2.0 and Whisper models for ASR tasks, optimizing performance with appropriate tokenizers.
• Utilized Torchaudio and librosa for efficient audio preprocessing, optimizing parameters to significantly reduce latency and improve transcription accuracy.
• Conducted extensive local testing with diverse audio samples and successfully deployed the transcription pipeline as a scalable web service.

Led the design and development of a complex payload, demonstrating leadership and technical mentorship.

• Successfully led the design and development efforts as Lead Payload Engineer for the Spaceport America Challenge.
• Mentored a team of 4 engineers as Payload Subsystem Leader, providing technical guidance and fostering skill development.

Completed a 10.2 km half marathon, showcasing dedication and physical endurance.

• Successfully completed the IIT Bombay half marathon (10.2 km category) organized by Aavhan, demonstrating commitment and discipline.

Actively participated in team sports, fostering teamwork and sportsmanship.

• Actively participated in Pluto Fiesta, the annual event of Hostel 9, playing Football and Basketball, showcasing strong teamwork and sportsmanship.

Developed an advanced predictive model and simulation tool for stock market analysis and forecasting, optimizing financial strategies.

Engineered a secure and scalable bank management application featuring Face ID authentication and comprehensive banking functionalities.

Achieved a top rank among 150,838 highly competitive candidates nationwide, demonstrating exceptional academic aptitude.

Secured 96.2% in Matriculation, demonstrating a strong academic foundation and consistent high performance.

Secured a top rank among 1.034 million candidates nationwide, reflecting strong foundational knowledge in science and mathematics.

Awarded a prestigious fellowship in the SA stream, recognizing outstanding talent and aptitude for research in basic sciences.

Achieved a high international ranking in a competitive science olympiad, showcasing strong scientific knowledge and problem-solving skills.