The following is an interview with Andrew Shpanch, Founder and CEO of CryptoMatch, about how the use of AI combined with zero-knowledge proof provides a foundation for trust and connections in the web3 space.
The crypto world is a place of distrust – and that must be fixed
As a first question, Andrew was asked what inspired him and his co-founder to create CryptoMatch.
“I got into web3 about a year and a half ago when I teamed up with Anastasia Moroz, my current fellow founder. As someone with web2 startup experience, I found web3 lacked trust. It was a challenge that needed to be addressed,” he said.
Andrew believes people are extremely cautious and cannot blindly trust each other in the crypto space because it is a Wild West place full of scammers and rife with frauds. According to him, there are a lot of blockchain projects to choose from, but it is difficult to locate trustworthy ones.
“We wanted to make a platform that would go beyond individual users and take into account not only what they say about themselves and what they aim for, but also what other industry players are saying. The idea was to consider as many factors as possible, such as user goals, industry feedback, and more, to ensure that users find the right network connections and reliable blockchain projects,” he explained.
After that, Andrew discussed the challenges associated with starting CryptoMatch. In his view, the most challenging part was gaining trust from first users and getting them to use the product.
“Luckily, Nastya, the other co-founder, had a big web3 network, so we used it to build the first audience and bring in industry pros. We formed an advisory board with experts like Stepan Gershuni and Vadim Koleoshkin, co-founder of Zerion, both of whom not only have a deep knowledge of the crypto world, but also a proven track record of success,” he said.
According to Andrew, CryptoMatch’s uniqueness is that it is free to use and speeds up business communication in the web3 era.
“We’re not the direct source of income, but right networking can help startups grow faster. Basically, we’re optimizing communication and finding the right people, two vital aspects of any business. The web3 space is in desperate need of trust-building mechanisms, and CryptoMatch is doing just that,” he said.
Why do we need AI matching bots?
Andrew then explained how he would explain CryptoMatch’s AI matching bot to a non-techie, as well as the benefits of using an AI matching bot over manual matching.
“By creating a profile, you give the AI algorithm the information to work with to find the best match for you. It looks for potential matches and notifies you when one is found.
Afterwards, you can accept or decline the match, and if it’s accepted, you set up a time to meet. At the end of the meeting, you give feedback to the algorithm so it can learn and improve its matching process. Simple as that!” Andrew explained.
On why it’s better than manual matching, Andrew pointed out that manual processes can only handle a limited amount of data and the best candidates are not always obvious. Also, it takes a lot more time.
“AI makes this process faster, using more parameters and collecting more feedback on matches, giving the user more options and a higher success rate in finding the right match. We’re planning to add a formula that takes into account both feedback from users and their ratings. The formula will be public, so everyone will be able to understand how it works,” he continued.
CryptoMatch’s security, upcoming features, and plans
Additionally, Andrew addressed security measures, upcoming features, and CryptoMatch technology and plans.
“We’ve learned how to tag searches and profiles. Right now, we’re using OpenAI embeddings to generate recommendations. For each match, the system displays the 10 most relevant options along with predicted feedback. The human element is still key at this point, and we use our staff to make final decisions,” he said.
As well, he pointed out that CryptoMatch is also protected from malicious attacks by API Telegram security measures, such as strict limits on requests per second (QPS).
“Authenticating our bot is just like signing in to Telegram – it’s pretty hard to hack it. And to top it all off, a 2FA option is always available as an added layer of security,” he said.
In terms of strategic plans, Andrew said CryptoMatch has almost completed its $1 million angel round. It intends to invest a portion of the funds in developing its matching engine so that it can work without human intervention. Also, the funds will be used for marketing and partnerships in the web3 space to connect people from industry-defining sectors. The platform currently has 15K users, but Andrew says it hopes to reach 100K by December. The lion’s share of these users are founders, C-level executives, investors, and digital professionals.
“As we move forward, now that we have a decent number of users on Telegram and Discord, we’re about to launch a web platform with upgraded features like crypto wallet authentication and on-chain matching. Also, we’re implementing a zero-knowledge proof (ZKP) system that displays user feedback anonymously, without revealing the author’s position or competence. This way, we’ll be able to get an accurate, unbiased image of each user, and match them accordingly,” Andrew added.
According to Andrew, CryptoMatch has even seen some unexpected uses, “We’ve had people using CryptoMatch to find “a startup girl” for instance, but not that often. We’re here to help people find their perfect business match – not their perfect soulmate. One day, maybe we’ll get into helping with dating, but not anytime soon,” he said.
AI looks at what people say about you as well as what you say
Also, Andrew talked about the steps being taken to ensure CryptoMatch‘s long-term success in order to stay ahead of the competition.
“In the web3 space, no similar solutions exist for solving trust problems. Other services connect people based on their social network data or their own descriptions, while we also take into account what other people think of them. Our main competitor is a warm introduction, which every single person asks for.
“It’d be awesome if other startups taught people how to network, so they could get targeted requests instead of just bringing sand to the beach like they do now. Still, we make sure to partner with some of the most respected firms to stay on top. With a quality product and efforts like this, I think we can keep growing and stand out,” Andrew said.
Web3 and AI have a bright future together
Finally, Andrew discussed how he views web3 and artificial intelligence evolving together in the future and what advice he’d give young entrepreneurs looking to break into the field.
“I think AI will be tightly integrated into web3 projects, and there will be a lot of unicorns at this intersection. With web3 technology, you can build secure and personalized services. Currently, the technologies are already intertwining. For example, AI can boost smart contract auditing processes or make them more customizable.
AI can also streamline DAO decision making. You know, DAOs spend a lot of time and resources validating and managing proposals, so AI can automate the process and make it easier.
Besides that, AI can help with crypto trading – perhaps the most desirable use case for an average user, which already works nicely. However, I’d caution new traders against taking AI forecasts at face value and blindly following them,” he warned.
Speaking about web3 entrepreneurs looking for a foothold in the industry, Andrew says it’s wise to start with forming a team with responsibilities split. “Even if you’re all passionate about the product, someone has to build it, someone has to get users, and someone has to raise money and that’s how you guys win,” he said.
“It’s challenging and expensive to find good headhunters. Our experience proves that a network developed in advance can help at the right time. To take one example, we secured a highly qualified CTO who was involved in multiple AI projects. And it’s not luck; networking helps us influence the right circumstances,” Andrew concluded.
