Interview du plateau de gestion d'actifs cryptographiques Fyde : comment utiliser l'IA pour créer des coffres-forts sur la chaîne afin d'accroître les rendements, réduire les risques et optimiser la liquidité ?
TechFlow SélectionTechFlow Sélection
Interview du plateau de gestion d'actifs cryptographiques Fyde : comment utiliser l'IA pour créer des coffres-forts sur la chaîne afin d'accroître les rendements, réduire les risques et optimiser la liquidité ?
L'IA et l'apprentissage automatique conviennent parfaitement à la surveillance, il est donc pertinent de les utiliser pour la détection préventive des menaces ou l'analyse du trafic réseau.
Dédié à des analyses Web3 approfondiesInterview : 1912212.eth, Foresight News
After experiencing a series of blow-ups in the previous cycle, some users have realized that many so-called asset management firms or protocols did not truly act in good faith to safeguard user funds. Some misappropriated funds for unauthorized lending, others suffered devastating losses from hacker attacks—users’ entrusted assets vanished overnight, leading to severe financial damage.
Truly reliable Web3 asset management platforms with unique advantages remain rare. The core mission—how to protect user funds while maximizing returns—remains unchanged. Recently, the crypto asset management platform Fyde sat down with Foresight News for an exclusive interview, discussing their views on industry trends, AI, and more.
Currently, most so-called AI-powered solutions are merely marketing gimmicks. However, Fyde extensively leverages AI for risk management and liquidity optimization, helping users earn better yields and compound faster. Additionally, backed by a strong team, Fyde plans to focus on consumer narrative tokens, restaking, and other emerging areas. Below is the full interview transcript.
From Crypto Hedge Fund to AI-Driven Asset Management
Foresight News: Some readers may not be familiar with what Fyde does. Could you briefly introduce the project?
Crypto users, including myself, often struggle to track where their tokens are held across different portfolios. They can't easily monitor which tokens are performing well or poorly, and frequently miss timely exits from profitable positions or fail to cut losing ones. Often, users don’t even have time to react—or claim airdrops. Their biggest question becomes: “Which rising category or narrative token should I buy or sell first?”
Fyde’s liquidity vaults solve this problem. We are an on-chain, AI-driven portfolio rebalancing and liquidity optimization engine. Users can deposit one of many different types of tokens, and our liquidity vault automatically allocates it across various tokens and narratives visible on our dApp.
By distributing holdings across a diversified range, users increase their chances of capturing major token rallies. Profits are locked in, while losses are minimized. AI and machine learning identify risks before they materialize, protecting users’ funds from volatility or isolated, extreme price movements. The vault continuously monitors and manages all risks—so users don’t have to. The goal of the vault is to help users consistently lock in gains, earn yield, and maintain liquidity, enabling faster accumulation of crypto holdings with lower volatility.
Foresight News: From what I know, Fyde’s team lineup is quite impressive, featuring former members from NASA, JPMorgan, Wintermute, and Synthetix. Why did you choose to enter the asset management space? Any stories to share?
Interestingly, we originally planned to become a cryptocurrency hedge fund. We started with four founding members, all with experience at large asset management and private equity firms, and deep backgrounds in quantitative research. That made sense—but during our development, we noticed that many people wanting to use our services were actually DAOs, protocols, and native crypto participants.
Their positions were highly concentrated, or they simply held large amounts of assets on-chain and wanted exposure to different risk profiles and emerging crypto narratives—but needed someone to manage this for them. We realized that if these users were willing to move their assets off-chain into illiquid crypto hedge funds, it clearly indicated a lack of high-quality on-chain solutions.
At the same time, we launched the LBS Blockchain Association and LBS DAO, and began searching for ways to manage our own assets. We found it extremely difficult to find anything that truly met our needs. So we decided to abandon the hedge fund idea and instead build a robust on-chain portfolio and liquidity management solution—one that could lock in gains and use AI for risk management on our behalf.
The Spark of AI Meets Crypto
Foresight News: AI is everywhere now, reshaping every industry. What’s your take on the convergence of AI and crypto?
We believe there are very few real use cases for AI in crypto today. Some genuinely interesting applications focus on risk reduction—which is exactly how we use it. But many AI products add little real value, or falsely claim to be AI when they’re just basic data analysis.
AI and machine learning excel at monitoring, so using them for preventive threat detection or network analysis makes perfect sense—these are currently the strongest applications we see.
Foresight News: Specifically, how does AI contribute to Fyde’s asset allocation, risk control, and market forecasting?
We use AI in two main areas: risk management and liquidity optimization. For risk management, we run machine learning over our network analysis. Machine learning excels at identifying patterns invisible to humans, allowing us to analyze relationships between on-chain transactions and token prices to quantify risk. For example, rapid wallet consolidation or large volumes of wash trading between wallets are red flags indicating potential issues—we become more cautious accordingly.
As another example, we’ve integrated Hypernative, which uses machine learning for proactive threat monitoring. They scan on-chain transactions to preemptively detect risks such as liquidity pool exploits, governance attacks, or hacks. We run their tools on tokens within our vaults and isolate any flagged assets—adding an extra layer of risk prevention and security.
These tools serve as safeguards against rug pulls, hacks, and similar threats, providing a layer of protection for user assets that most individuals cannot access on their own.
Finally, we're exploring AI for liquidity optimization. We create advanced simulation environments to test and optimize our token’s liquidity under various market conditions—enabling us to maximize DEX liquidity per dollar and make real-time dynamic adjustments based on machine learning outputs.
Foresight News: How do you train and optimize Fyde’s AI models?
We use AI in several different ways. In network analysis, we apply a subset of AI known as graph machine learning. Here, we build a network graph incorporating numerous on-chain parameters. These parameters help predict information about wallets (e.g., whether linked to market makers) and transaction flows (e.g., predicting traffic into Uniswap V3). This helps us identify centralization risks, wash trading patterns, and more.
To ensure our tokens achieve super liquidity, we run agent-based simulation modeling to determine optimal DEX liquidity structures. These agents simulate DeFi participants’ trading behaviors, modeling ideal arbitrage sizes, trader behavior patterns, etc.
All these inputs go into our simulation engine, which uses methods like Monte Carlo analysis—not only to uncover network patterns invisible to the human eye but also to determine the best way to provide on-chain liquidity.
What Sets Us Apart?
Foresight News: What are the unique aspects of Web3 asset management compared to traditional asset management?
Web3 asset management offers far more flexibility than its traditional counterpart. Of course, in Web3 you have classic strategies like buy-and-hold, yield farming, delta-neutral trading, etc.—but with significantly higher volatility.
While there's overlap, there's also massive room for innovation—especially around portfolio rebalancing and trade execution. In traditional asset management, you have dedicated execution traders. At a bank managing portfolios, there might be 50 such people whose job is to execute trades at optimal prices, each earning around $200K annually—totaling $10M per year in costs. Instead, we incentivize the entire ecosystem to perform these tasks for us.
Now, everyone in crypto is a potential execution trader—not just 50 individuals. By offering market-driven incentives, we drastically reduce costs—far below that $10M mark. This democratization of tasks is one of the coolest things about Web3.
Foresight News: In the competitive Web3 asset management landscape, what are Fyde’s key competitive advantages?
When people look at DeFi, they usually focus on yield generation. But unfortunately, if your underlying assets drop 95% during a bear market, a 20% yield means nothing. Our protocol aims to help crypto users consistently lock in gains while protecting against rugs, hacks, and other risks—so their portfolios grow steadily over time, with less downside impact. This allows faster compounding and generational wealth accumulation.
We use AI and machine learning to better manage user risk, shielding them from rug pulls, severe negative price actions, and more. Of course, we employ a whitelist system ensuring only pre-vetted assets enter our vaults. Our whitelist filters aim to eliminate scams and pump-and-dump schemes.
Additionally, users can access the platform using any whitelisted token. Most projects today only accept ETH, wBTC, or stablecoins, but with us, users can deposit any whitelisted token—making this approach accessible to more people looking to grow and protect their crypto portfolios.
Trends and Challenges
Foresight News: What technologies or trends in the Web3 asset management space are worth watching?
L2s and how to effectively utilize them will be crucial for the future of Web3 asset management. Running portfolios efficiently and managing risk requires executing a large volume of transactions. Unfortunately, doing this on the Ethereum mainnet is prohibitively expensive. Much of this work, however, can be done on L2s.
For instance, we run numerous machine learning and AI models, each requiring thousands of iterations. Imagine running this on the mainnet with $20 gas fees per transaction—we’d never be able to do what we need. On L2s, we pay just $0.02 per transaction, reducing our costs by 99.9%.
Foresight News: What industry challenges has Fyde faced during its development? How have you addressed them?
Our biggest industry challenge has been regulation. Currently, regulatory clarity in crypto—especially DeFi—is extremely low. But we believe you shouldn’t stop building just because of regulatory uncertainty. That said, we’ve taken all necessary steps to ensure our platform complies with current regulations.
We’ve completed all legal formalities, obtained legal opinions, properly structured our company, and more. With this foundation, we’re prepared to keep building, ensuring both our platform and users are protected. As regulations evolve, we maintain active dialogue with legal counsel to stay up to date.
Foresight News: Can you share Fyde’s future development roadmap?
Absolutely. We plan to do many exciting things on L2s. Our current liquidity vaults focus on risk protection, but going forward, we aim to expand into consumer narrative tokens, restaking, and other vault types that interest users. We’re particularly excited about building a vault for MEME tokens—allowing users to invest without fear of rugs. We’ll use AI to identify the most promising MEME tokens for purchase. We also plan to leverage MEV to enhance the yields we currently offer in our liquidity vaults.
Next year, our goal is to make our token as liquid as possible, further incentivizing user adoption and helping us expand deeper into B2B markets.
Foresight News: What advice would you give to investors or users interested in entering Web3 asset management?
For users, my advice is to check who’s on the core team before using any Web3 asset management product. Many developers have never worked at asset management firms or understand risk management at all. These aren’t the people you should trust with your money and time—and that’s precisely why so many blow-ups happen in crypto. Choose products with solid teams and meaningful designs.
For builders, I’d say: truly understand market trends. It’s easy to get caught up in flashy tech and shiny new things, but ultimately, technology must be built with a clear purpose.
Bienvenue dans la communauté officielle TechFlow
Groupe Telegram :https://t.me/TechFlowDaily
Compte Twitter officiel :https://x.com/TechFlowPost
Compte Twitter anglais :https://x.com/BlockFlow_News
Foresight News![Analyse approfondie de Trade[XYZ] : comment 92 marchés et 98 % du volume de trading HIP-3 ont-ils été établis ?](https://upload.techflowpost.com/upload/images/20260716/20260716061117965147.jpeg?x-oss-process=image/resize,p_50/quality,q_80)
