Waterloo students win nearly $100k USD at ETHDenver 2025

Seventeen students from the University of Waterloo and Wilfred Laurier University dominated ETHDenver, held from February 23 to March 2, 2025.

ETHDenver is the world’s largest and longest-running Ethereum event. Since 2018, it has hosted blockchain-focused events, including panel discussions, workshops, networking sessions, and boot camps. Their main event is #BUIDLathon, a hackathon where attendees can team up to develop their own blockchain projects.

During #BUIDLathon, the teams must submit their projects to five tracks: Infrastructure + Scalability, Identity + Privacy + Security, DeFi + NFTs + Gaming, DAOs + Communities, and Impact + Public Good. During the premier round, the teams will present their projects to a panel of 50 community judges, who will select three winners from each track. The top 15 finalists are invited to ETHDenver’s closing ceremony and pitch to a panel of five celebrity judges. Each judge will select one winner from each track.

In addition, #BUIDLathon’s participants can submit their projects to any sponsor-related tracks and receive prizes called bounties. These sponsors include major blockchain companies like PolkaDot, Okto, and Zircuit.

Some of #BUIDLathon attendees included members of the University of Waterloo’s Blockchain Club: Fahim Ahmed, Ishan Baliyan, Manuel Stefan Christopher, Krish Chopra, Cole Dermott, Xavier D’Mello, Suyog Joshi, Joshua Kim, Nelson Lai, Eric Liu, Alex Lu, Pravesh Mansharamani, Daniyal Mohammed, Sailesh Polavarapu, Andrew Tretyakov, William Wang and Dasha Yefymenko.

“ETHDenver was just a great opportunity for us,” says Computer Science student Dasha Yefymenko. “First, I get to brainstorm and bounce ideas around and build something that I can potentially take to the next level, like creating a startup. There’s lots of networking and mentorship. If you win, you might get a financial prize, another avenue of help. Also, hackathons are awesome, because they’re hosted around the globe. So, you can work on something you're passionate about and see the world, all at the same time.”

The students competed against other juggernaut universities including UC Berkeley, New York University and University of Pennsylvania. Yet, Waterloo’s Blockchain Club was one of the top competitors, collectively winning $84,500 USD (approximately $120,100 CAD) out of a prize pool of $1,029,833 USD.

Eigen Games, a competition organized by EigenLayer and CollegeDAO, was held in partnership with ETHDenver. Its finals took place at ETHDenver’s venue. Two teams — EigenBets and FaceBuddy — collectively won $10,500 USD at Eigen Games, bringing the students’ total win to $95,000 USD (approximately $135,426 CAD)

The students credit their win to UWaterloo’s and Laurier’s in-depth courses and co-op system, which exposes students to modern-day and critical challenges and solutions.

“Having six co-ops kind of builds a builder mentality around Waterloo students,” says Xavier D’Mello, a computer science student at Wilfred Laurier University. “We have more practical building experience than other schools who may focus more on theoretical business.” 

The students didn’t just win over the judges, but also fellow ETHDenver attendees and hackers. Two Waterloo Blockchain’s projects — FaceBuddy and zkHotdog — respectively ranked first and second place in community voting.

Below are descriptions of the students’ award-winning projects.

Students: Fahim Ahmed and Dasha Yefymenko (Computer Science)

Prizes: Top three finalist for DEFI, NFTS + Gaming track ($5,000) and Build an AI-powered app on Base ($5,000)

We live in a world where AI is evolving at such a rapid pace, that we don’t know if an image, video or text message is truly real. “There's a concerning number of bots online, which influence national politics or people's individual view of the world,” explains Fahim Ahmed, who created Bot or Not with his classmate Dasha Yefymenko. “A lot of these bots end up spreading propaganda or false information. Humans need to figure out who is a bot or who's not.”

Bot or Not is an amalgamation of the Turing Test, a classic philosophical concept and Among Us, a staple pandemic game. To join, players must pay $10 each into the prize pool. Onwards, they are brought to a group chat, where they must converse and identify which player has been randomly selected to have their AI chat on their behalf. At the end of the game, the prizes distributed to human players who correctly identify the AI agent. However, if the AI went undetected, the entire prize pool goes to the user who used prompt engineering to make their AI agent act human-like.

With critical applications to the real world, the team plans to collaborate with Base, a sponsor of #BUIDLathon bounties, to bring this game to the app store.       

Students: Ishan Baliyan, Krish Chopra, Christopher Simanjuntak (Statistics and Combinatorics & Optimization), and Daniyal Mohammed (Data Science)

Prize: Hacken AI Agent Track Winner ($3,000)

One of blockchain’s biggest game-changers is smart contracts, computer programs that automatically execute an agreement once the predetermined conditions are met. These programs do not need any intermediaries like a bank or lawyer, making these agreements trustworthy and efficient. 

However, smart contracts are still suspectable to vulnerabilities. “Let’s say a developer types a specific line,” explains Christopher Simanjuntak. “While the code might be syntactically correct, it could have a loophole such that hackers could exploit the system, like draining the whole wallet or financial system.”

Although one could manually audit these codes, it can be time-consuming and expensive.

Christopher collaborated with Ishan Baliyan, Krish Chopra, and Daniyal Mohammed to build Brokechain. This AI agent can scan a contract’s code to detect any security loopholes, vulnerabilities and inefficiencies. It also creates a detailed security report on risk levels and improvements. By automating this process, people can deploy contracts faster and safer.

One of the bounties that the team applied for and won was Hacken, a blockchain-based cybersecurity firm. Hacken may implement this project to improve their current workflow, showcasing the cutting-edge nature of our students’ work.

Students: Suyog Joshi (Mathematical Physics), Joshua Duho Kim (Computer Science), Eric Liu (Waterloo & Laurier’s Math and Business), and Alex Lu (Computer Science)
Prizes: Build an AI-powered app on Base ($5,000), Okto Consumer Dapp First Place ($3,000), Building a Decentralized Content Management System using Web3 and EthStorage ($250) and Building a Decentralized Social Network using Web3 and EthStorage ($250)

Cyworld was the premier social media platform in South Korea during the mid to late 2000s. Its platform was a unique blend of Sims and Myspace, where users design their own rooms and share blog posts. Although the platform officially shut down in 2019, a group of Waterloo Math students — Suyog Joshi, Joshua Duho Kim, Eric Liu and Alex Lu — have reimagined its core concept with a touch of blockchain, AI, and 3D elements with their project, Cyspace.

Social media has always been criticized for fostering surface-level relationships among users. Cyspace, however, introduces an AI agent that scans users’ posts to generate personalized quizzes. Users who successfully complete these quizzes are rewarded with a non-fungible token (NFT), a blockchain-based asset, that proves their friendship and can be used to unlock deeper interactions. This concept adds an interactive and meaningful dimension to the social media landscape beyond superficial likes and comments.

Students: Cole Dermott (Waterloo & Laurier’s Math and Business) and Pravesh Mansharamani (Math)
Prize: EigenLayer AI track First place ($7,000)

Cole Dermott and Pravesh Mansharamani built EigenBets, a prediction market platform that allows users to place bets on either side of an outcome, like will Bruno Mars tweet 6,000 times this month? Their system uses AI agents to scan data to verify the outcome, eliminating human bias and intervention. Users can also set which sources of evidence the AI agent must use, such as relying on verified Twitter accounts, and time for the market to end. 

“There’s a theory that when there’s money on the line, people are more likely to bet on the real outcome or the truth of the matter,” says Cole.

Pravesh explains that “most market platforms like Wealthsimple are based on investing. EigenBets is built around conviction in certain trades or outcomes that are happening right now. It essentially allows people to put their money where their mouth is.”

Most betting markets employ order books, meaning that someone can only buy a share if someone else is selling the equivalent amount. However, EigenBets uses automated market making which doesn’t restrict trading, leading to more liquidity in the market. This unique feature allows more seamless trading and a better user experience, compared to other market platforms.

EigenBets also won prizes amounting to $9,000 USD at Eigen Games.

Students: Xavier D’Mello and William Wang (Computer Science)

Prizes: Top three finalist ($5,000) and Celebrity Judge Pick ($1,000) for IMPACT + PUBLIC Goods track, Build an AI-powered app on Base ($5,000), Uniswap’s DeFi Innovation on Unichain Second Prize ($1,500), Humanity Protocol’s Web Application for Verifiable Credentials Second Prize ($5,000), Consumer Crypto Prize - Coinbase Developer Platform ($2,000) and Walrus Tusked Champion Third Place ($1,500)

Second-time champions, Xavier D’Mello and William Wang, created FaceBuddy. This app allows users to exchange social media information or money through facial recognition. Instead of awkwardly trying to search someone’s account and accept their follower’s request, you can simply scan their face.

“We’re using this library for face recognition called Face API DOT JS,” says William, a student at the University of Waterloo. “It takes an image of your face and creates a hash out of it, which means it’s able to get the distances between your facial features such as your eyes, nose and mouth. You might look similar to someone, but mathematically there will be variations between your facial features.”

The app aims to “bank the unbanked.” Around 1.7 billion people worldwide lack access to a bank, making it harder to accept and receive payments. However, most of them have a smartphone.

“Imagine you’re a foreign worker,” says Xavier. “You’re moving to a richer country, but you don’t have anything with you. How do you get paid? Well, something that everyone has is a face. It doesn’t discriminate. It doesn’t cost money.”

It can also benefit avid travellers since every country has different banking regulations and currencies. William recalls how difficult it was to pay someone back at a hackathon in America since American cash apps like Venmo aren’t available in Canada.

Xavier and William were selected as the Top Three Finalist and Celebrity Judge Pick, just like last year. They also won various bounty prizes at #BUIDLathon and EigenGames.  

Student: Sailesh Polavarapu (Computer Science)

Prizes: First Prize for Hedera AI and Agents Challenge ($15,000) and ZK Proofs with New Technology ($5,000)

Waterloo student Sailesh Polavarapu co-created InfoFi with Sebastian Liu, a recent graduate from the University of Washington. This project intersects AI, crypto and privacy. 

Their database allows data vendors to upload their datasets, and customers to browse through various datasets to train their AI models.

“The problem is that data vendors might have datasets they’d want to sell, but there’s a huge risk of their set being leaked and people reselling it for a much lower value,” says Sailesh. “We need to provide a way to protect vendors from people stealing their data sets.”

InfoFi does not reveal the content of the vendor’s dataset to a potential customer. Instead, it asks the customer for their preferred accuracy rate and a validation dataset, a type of dataset used for evaluating and fine-tuning a model before final testing. Onwards, InfoFi will train a vendor’s model and then run it against the validation dataset to see if it meets or exceeds the customer’s accuracy threshold. If it does, the customer pays for and receives the model. If it doesn’t, the customer won’t have to pay.”

InfoFi employs zero-knowledge (ZK) proof, a blockchain protocol that can prove something is true, without revealing its details. For example, authenticating a vote without showing the voter’s selection. In this case, InfoFi uses ZK proof to show that the inference on the vendor’s dataset matched the accuracy threshold.

Student: Nelson Lai (Actuarial Science)

Prizes: Flow’s AI Agents’ Fourth place ($2,000) and zkSync’s Best Web3 Onboarding UX using ZKsync Smart Sign-On SDK’s Third Place ($2,000)

Cryptocurrency has become the hottest financial tool in recent years. However, its novelty can scare newcomers away, especially those unfamiliar with blockchain. Enter AI chatbot Personal Finance Girlfriend, created by Nelson Lai, that can explain and handle cryptocurrency to you in a loving way.

Personal Finance Girlfriend allows users to create tokens, NFTs and smart contracts using plain language. In essence, they don’t need any coding knowledge. Personal Finance Girlfriend can also transfer tokens and manage budgets. This innovative app can encourage and support users throughout their cryptocurrency journey, just like their significant other. 

Student: Andrew Tretyakov (Mathematical Finance)

Prizes: Top three finalist for Identity, Privacy + Security track ($5,000) and ZK Proofs with New Technology ($5,000)

Waterloo student Andrew Tretyakov created zkHotdog, an app that can determine the size of any object in real life or from a photo. It can also create a cryptographic proof of its size. During his presentation’s demo, he used the app to prove the size of a hot dog. The app was built using ZK proofs, hence the name.

Andrew was inspired by his research into ZK proofs and an inside joke involving using ZK proof to prove the existence of someone’s girlfriend. Despite facing some challenges with the presentation, he ended up winning the #BUIDLathon’s Identity, Privacy and Security Track, alongside zkVerify Foundation’s bounty prize. He hopes to expand the app by implementing more real-world usage, such as in civil engineering or the modern dating landscape.