From Classroom to Consensus: Humber BIA Student Wins $10k at Global Hackathon
HLOs: 
How AI, Web3, and Grit Drove Shyamal's Second-Place Win at the Aptos x EasyA Hackathon
In a remarkable demonstration of curiosity, collaboration, and innovation, Business Insights and Analytics (BIA) student Shyamal Bhatt secured second place and a $10,000 prize at the EasyA x Aptos Hackathon held during Consensus 2025 in Toronto.
What makes this achievement extraordinary is not just the recognition, but the journey that led there. Just five days before the event, Shyamal and his teammates were strangers with no prior experience in Web3 technologies. Within 72 hours, they had conceptualized, developed, and demoed AptosAI Guard—an AI-powered wallet risk detection tool built on the Aptos blockchain. Their innovation was selected among the top three winners out of 36 teams competing on the Aptos challenge track.
A Real-World Problem Meets an AI-Driven Solution
The idea emerged from a simple but critical insight: sending tokens on blockchain platforms is often a leap of faith. Could users be alerted before transferring funds to a potentially suspicious wallet?
Their answer: Yes, using machine learning.
AptosAI Guard fetches a recipient’s transaction history, analyzes behavioural patterns, and generates a real-time risk score, all before the transaction is signed. The tool helps prevent scams and risky transfers in real time, making blockchain usage safer for everyday users.
“We built the entire project—from data extraction to machine learning modelling to frontend design—within just three days, and we used a Random Forest algorithm to generate risk scores between 0 to 5 based on wallet age, diversity, NFT holdings, and historical behaviour.”
- Shyamal Bhatt
The solution features a backend built in Flask, a React-based frontend, and integration with APIs from Aptos and CoinGecko. The team open-sourced the project on GitHub and is now working on turning it into a browser extension, a minimum viable product (MVP) they hope to improve and pitch for further development funding.
Meet The Team: Diverse Talents, Shared Vision
The team was a multicultural mix of students and early-career professionals, many meeting for the first time at the hackathon:
- Shyamal Bhatt – Machine Learning Lead (Humber Polytechnic, BIA Program
- Jessica Sumin Lee – Backend Developer (Software Developer, Bell)
- Felipe García López – Frontend Engineer (Software Engineer, Orium)
- Nolan Druid – Project Manager (Computer Science, Carleton University)
Despite their different backgrounds, what united them was a shared commitment to learning, problem-solving, and building something meaningful, fast.
“We knew nothing about Web3. but we believed in the problem, leaned into each other's strengths, and kept things simple."
- Nolan Druid
“This project reminded us that simplicity often lies on the other side of complexity."
- Shyamal Bhatt
The role of Humber DAC: Where Curiosity Meets Opportunity
This opportunity came through the Humber Data Analytics Club (DAC), where Shyamal is an active member. It was shared with the community after Alexandra Joanna Baker from EasyA reached out directly.
Shyamal’s win reflects Humber’s commitment to applied, industry-focused learning and highlights the impact of student-led initiatives like DAC, which empower students to explore emerging technologies, build portfolios, and make a global impact.
“I wouldn’t have even known about this event if it weren’t for the DAC’s post. The club isn’t just a space for analytics—it’s a gateway to real-world challenges and global opportunities.”
- Shyamal Bhatt
The team is now focused on:
- Building a user-ready browser extension
- Enhancing the machine learning model with multivariate LSTM algorithms
- Integrating scam blacklists and user feedback loops
- Exploring funding opportunities with Aptos to expand the platform
They’ve already shared a project walkthrough via Loom and open-sourced their code on GitHub for others to learn and build upon.
Apply to this program