top of page
Writer's pictureDecasonic

AI Meets DeFi

Can AIFi Turn Anyone into a Successful Quant Trader? By Kathy Tong, Quantitative Venture Investor at Decasonic


Early DeFi focused on moving liquidity with protocols like Uniswap and AAVE. Now, DeFi emphasizes broader accessibility, enhanced UI/UX, reduced fees, and expanded liquidity options. Recently, we’ve seen a resurgence in DeFi with the rise of platforms such as Pendle, Hyperliquid, Fluid DEX, Ethena among many new innovations. These innovations now enable a new method of maximizing smart collateral and smart debt, enabling efficient liquidity management and deep liquidity in perp trading and expanded liquidity options. 


At the same time, AI is upgrading DeFi, driving the emergence of "AIFi," where AI agents simplify and automate complex financial tasks in decentralized finance. This convergence of AI and DeFi promises to break down barriers, enabling more users to harness the power of decentralized finance through intelligent, adaptive systems that enhance decision-making and operational efficiency.


AI agents, integrated with blockchain technologies, represent the future of decentralized finance by addressing operational inefficiencies and lowering barriers to entry. This article explores the role of AI agents in DeFi through case studies of leading platforms: Almanak, Mode Network and Spectral Labs.


Resurgence of DeFi


DeFi has seen substantial growth in recent years, with Ethereum currently leading in TVL at $191b, and Solana leading with active addresses at 4.2 million users. However, DeFi’s trader  and user base remains relatively small compared to the billions served by traditional financial systems. The reasons for this disparity may include complexity, fragmentation across blockchains, and security concerns. Users must navigate fragmented ecosystems with varying protocols, creating operational inefficiencies and liquidity silos. Smart contract vulnerabilities further deter adoption by increasing risks of financial loss. These challenges highlight the need for automated, intelligent solutions to streamline DeFi’s operations and enhance user accessibility.


TVL locked in all DeFi YTD


DeFi YTD

Net flows in all chains YTD:

Net flows in all chains YTD

Introduction of AI Agents in DeFi


AI agents are autonomous software applications designed to analyze data, make decisions, and execute transactions within the context of a DeFi ecosystem. Unlike human users who need to interact with protocols directly, AI agents can seamlessly optimize tasks such as liquidity management, cross-chain trading, and portfolio rebalancing. At Decasonic, I’ve talked to many innovators building on this frontier of AI and DeFi, looking to solve these challenges and expand the adoption of decentralized finance.  


The emergence of these agents has been driven by advancements in machine learning and quantitative reasoning. For example, recent AI models like OpenAI’s GPT-4 and Anthropic’s Claude have demonstrated remarkable capabilities in quantitative problem-solving, setting the stage for more sophisticated AI-driven systems. These agents are positioned to reduce DeFi’s complexity by automating decision-making processes and providing intuitive interfaces for users.


Platforms like Almanak, Mode Network, and Spectral Labs are leveraging these technologies to enable users to create, train, and deploy AI agents that can operate across multiple chains and protocols, enhancing efficiency and accessibility in DeFi.


Case Study 1: Mode Network


Mode Network has positioned itself as the AI x DeFi layer 2 within the Optimism Superchain, focusing on creating an ecosystem that simplifies DeFi interactions by being a platform for AI agents to transact seamlessly on-chain. 


Currently, there are 120+ agents live on Mode with leading agents from Giza, OLAS, Amplify and Brian. Mode’s AI Agent App Store allows users to discover and deploy agents designed for specific tasks across different chains and protocols. 


Mode Network

Below, let’s walk through how to use Brian’s intent based chat platform to automate transactions. Brian’s app can be accessed here. Users are able to ask Brian, send transactions, deploy contracts and search data. 


To send a transaction, users simple: 

  1. Connect Wallet

  2. Through a chat-style interface, users specify their transaction or financial objective (e.g. swap 1 USDC for ETH) 

  3. Users approve the transactions in their wallets

View on BaseScan the transaction’s progress


Brian

Mode also has a subnet on Bittensor ecosystem - Synth subnet. Synth is a data layer integrated into Mode—enhances the forecasting capabilities of its agents and improves accuracy of financial data sets and predictions. For example, agents can predict optimal liquidity ranges or price options more accurately, ensuring higher efficiency in managing financial resources.


Case Study 2: Spectral Labs


Spectral Labs specializes in deploying AI agents for high-frequency trading within DeFi. Its Syntax platform enables users to permissionlessly create agents that autonomously trade futures on platforms like Hyperliquid. These agents are particularly effective in executing complex strategies such as perpetual trading, liquidity provision, and risk management.


Spectral Labs

Steps to create an AI agent on Spectral Labs:

  1. Connect Wallet: Similar to Brian, users start by linking their wallets to the Spectral Syntax app. 

  2. Define Agent Parameters: Users specify their agent’s name, ticker symbol, and the amount of tokens to allocate (ranging from $100,000 to $1 million, for a max of 1% total FDV).

  3. Train the Agent: Through a chat interface, users train the agent on desired trading behaviors, risk tolerance, and strategic goals. This process is similar to customizing a language model like GPT for specific tasks.

  4. Deploy the Agent: Once training is complete, the agent is deployed to autonomously run the strategies, and other users are able to also buy-in. When the agent reaches a certain market cap, it is deployed onto Uniswap. 


$SPEC’s tokenomics follow a dual-token model similar to Virtuals Protocol ($VIRTUAL) and other AI agent launchpads, where users use SPEC to swap for $AGENTCOIN and $SPEC provide the native liquidity needed in the agent-specific LP pool, prior to reaching the milestone needed for Uniswap deployment. 


How agent creation works:


$SPEC

Creating new agent UI fields: 

Creating new agent UI fields:

Agents created on Syntax ecosystem are designed to deploy a variety of AI-driven DeFi strategies, with a significant focus on perpetual trading on Hyperliquid. For instance, $YUKI, an AI agent with a current market cap of $270,000, aims to embody the spirit of calculated risk-taking and seasonal investment strategies, with a focus on secure asset storage. $YUKI conducts perp trading strategies on Hyperliquid with their recent trade history and current balance. 


$YUKI

Similarly, CELINE YEE ($CYA), with a current market cap of $516,000, aims to deploy a high-frequency trading strategies with a focus on high Sharpe ratios and directional risk-taking. 


Below are the recent trade history and current positions of $CYA on Hyperliquid. 


 $CYA

Case Study 3: Almanak


Almanak is a pioneering platform in the AIFi space, focusing on democratizing quantitative trading through AI agents. The platform has developed a system where users can create AI agents capable of analyzing markets, identifying opportunities, and executing strategies autonomously.


Almanak agents operate on what the platform describes as a "next-generation DeFi infrastructure," which minimizes human intervention. Users can build custom strategies or adopt existing ones shared by the community. This approach empowers individuals without technical expertise to engage in complex trading strategies previously dominated by institutional actors.


Key Features of Almanak AI Agents:

  • Analyze market trends and execute trades autonomously.

  • Provide customizable templates for strategy development.

  • Reduce operational inefficiencies by automating liquidity management.


Almanak’s long-term goal is to make AI agents the backbone of DeFi trading, enabling a more resilient and efficient ecosystem. The platform has also implemented robust testing environments to ensure the reliability of its agents, addressing security concerns often associated with DeFi protocols.


Future of AIFi


The integration of AI agents into DeFi is poised to redefine the financial landscape. Over the next decade, several trends are expected to shape the AIFi ecosystem:


  1. Widespread Adoption of Agent-Driven Interfaces: Traditional DeFi frontends will be replaced by intuitive agentic interfaces that automate complex financial interactions, across the various blockchains today.  This shift towards multi-chain accessibility will lower the barrier to entry for new users, driving mainstream adoption.

  2. Enhanced Intelligence and Forecasting: Platforms like Mode Network and Spectral Labs are already integrating advanced data layers to improve the predictive capabilities of their agents. This trend will continue, enabling agents to adapt to rapidly changing market conditions.

  3. Scalability Across Chains and Protocols: The ability of AI agents to operate seamlessly across multiple chains will enhance liquidity aggregation and reduce inefficiencies, making DeFi more accessible and efficient.

  4. Collaborative Ecosystems: Platforms like OLAS will play a critical role in fostering collaboration and standardization within the AIFi space, ensuring that innovations are shared across the ecosystem.


Initiatives like the AIFi alliance aims to revolutionize DeFi by integrating AI to create intelligent, autonomous financial systems that enhance efficiency, accessibility, and scalability.


Conclusion


The convergence of AI and DeFi represents a transformative moment in the evolution of financial systems. Recently, ecosystems like Aerodrome ($AERO), Hyperliquid ($HYPE), and Sui ($SUI) have sparked a resurgence in DeFi, demonstrating innovative approaches to liquidity management and trading efficiency. These platforms are not only advancing the core capabilities of decentralized finance but also laying the groundwork for a new renaissance of innovation. We expected an influx of AI-driven projects building on these ecosystems. By automating complex tasks and reducing barriers to entry, AIFi has the potential to democratize access to financial services, making them more inclusive and efficient.


Platforms such as Almanak, Mode Network and Spectral Labs among others are at the forefront of this transformation, providing the tools and infrastructure necessary for users to leverage AI agents in DeFi. As these platforms continue to innovate, they will shape the future of finance, enabling a seamless, intelligent, and adaptive ecosystem.

The journey of AIFi has only just begun, but its potential to revolutionize decentralized finance is already evident. By embracing this convergence, we can build a financial system that is not only decentralized but also intelligent and resilient, paving the way for the next era of global financial innovation.


The content of this material is strictly for informational and educational purposes only. It is not intended to constitute investment advice, nor should it be considered a recommendation or a solicitation to buy, sell, or hold any asset. Decasonic does not endorse investments in any specific tokens, and nothing in these blog posts should be construed as legal, tax, or financial advice. Please consult with a qualified professional advisor before making any financial decisions. Decasonic provides no warranties, whether expressed or implied, on the content provided in these blog posts, including its accuracy, completeness, or correctness. The opinions expressed here are those of the authors and do not necessarily reflect the views of Decasonic. Please note that Decasonic may hold a position in some of the tokens mentioned, including Virtuals. Decasonic is not liable for any errors or omissions in the content of this material or for any actions taken based on the information provided herein.



bottom of page