2nd Prize - BRICS International Vocational Skills Competition 2024
BRICS Skills Competition, Data Analysis and Visualization / China / 2024
I had the opportunity to represent India and Sister Nivedita University at the BRICS Skills Competition 2024 in China, competing in the offline Data Analysis and Visualization category. Among 178 international participants, I secured 2nd Prize, a result that reflects both technical preparation and performance under high-pressure, real-time evaluation.
This experience remains one of the most important milestones in my journey because it combined international competition, rigorous problem solving, and national representation on a global stage. It was also deeply motivating to see our mentoring ecosystem recognized, with Dr. Indranil Sarkar receiving acknowledgment as an Excellent Guidance Expert, along with continued guidance from Dr. Anirban Mitra and Dr. Swarup Kr Ghosh.
The achievement was not individual in spirit. It belongs to my university, the communities that supported me, and the people who trusted that we could deliver at international standards. The outcome reinforced my long-term commitment to building high-impact AI systems with strong technical depth and practical relevance.
3rd Place - ICDMAI 2025 International Offline Hackathon
International Conference on Data Management, Analytics and Innovation (ICDMAI) / Kolkata / January 2025
Our team from Sister Nivedita University, with Sohoomlal Banerjee and Sougata Sadhukhan, advanced from more than 1000 registered teams into a highly selective final cohort of 40 teams and ultimately secured 3rd place. The competition brought together participants from leading institutions including Jadavpur University, St. Xavier's College, IIIT Kalyani, and others, making the final ranking especially meaningful.
We built CredEase, a machine learning driven web platform designed to support street vendors through responsible micro-credit assessment. The system estimated optimal loan amounts and loan terms, produced adaptive interest profiles, and incorporated risk and creditworthiness analysis to improve fairness in financial decisions. To improve accessibility, we integrated LLaMA 3.3 Spectral based multilingual conversational support so users could receive guidance in their preferred language.
This project was shaped by strong mentorship and institutional support from Dr. Swarup Kr Ghosh, Dr. Anirban Mitra, and several mentors, jury members, and organizers who contributed throughout the event. The outcome validated our focus on socially relevant AI and demonstrated how applied ML can be translated into practical, user-facing financial tools.
1st Prize - ML Mania at PRAGATI 2024
MCKV Institute of Engineering / PRAGATI 2024
At ML Mania, hosted as part of PRAGATI 2024 at MCKV Institute of Engineering, our team with Sohoomlal Banerjee and me secured 1st prize after a high-intensity round of model design, experimentation, and technical presentation. The event emphasized practical machine learning depth and clear problem-solving under time constraints, and this win reflected the strength of our collaboration from ideation through final delivery.
Beyond the result, this milestone stood out because of the encouragement we received from peers, mentors, and the wider student research community. The experience reinforced our commitment to keep building systems that are technically rigorous, impact-oriented, and presentation-ready for real-world evaluation settings.
1st Place - Intel OneAPI Hackathon 2024
Intel OneAPI Hackathon / 2024
At Intel OneAPI Hackathon 2024, we secured 1st place with an Assistive System for Blind People focused on two high-impact tasks, obstacle awareness and currency recognition. The project combined real-time computer vision with practical usability so that visually impaired users could receive actionable feedback in day-to-day scenarios.
The system integrated YOLOv9 for object detection and MiDaS for depth estimation to identify nearby obstacles in real time, while a custom ResNet50 based currency reader recognized Indian note denominations and enabled audio-guided interaction. Our design objective was not only technical accuracy, but also confidence, accessibility, and reliability in real-world usage.
The project remains a strong example of purpose-driven engineering, where architecture decisions were guided by usability and social impact rather than benchmark performance alone.
1st Place (Tie) - SAP Inside Track Kolkata ICOE Hackathon 2024
SAP Inside Track Kolkata ICOE Hackathon / 2024
Our team tied for 1st place with two other teams at the SAP Inside Track Kolkata ICOE Hackathon, standing out in a field of 400 plus teams. Working with Archisman Ray and Sohoomlal Banerjee, I contributed to the design and implementation of an explainable automated essay grading system that balanced scoring quality with interpretability.
The solution used Bidirectional LSTM based modeling, including an attention-enhanced variant, and combined model outputs through ensemble averaging for stronger and more consistent grading behavior. Beyond leaderboard performance, the central contribution was explainability, making the scoring process more transparent and educationally meaningful for end users.
This win highlighted the value of collaborative research engineering under tight timelines, and reinforced my interest in building AI systems that are both high-performing and accountable.
