What are Intent Based Architectures?
DeFi puts complex financial products into the hands of users and protocols, enabling them to execute simple transactions like a swap - to more complex strategies like a delta neutral stablecoin based on staking yield and funding rate arbitrage (as was recently done by the USDe developer Ethena).
The problem for many users is that there exists no perfect protocol or tool to help them execute their specific strategy or workflow. Some users may take to Twitter to ask “Is anyone building the thing I need?”. The very brave may even try to manually cobble together their strategy by clicking buttons and signing Tx’s across multiple protocols; monitoring everything in a spreadsheet. This is the manual, transactional approach to DeFi. It’s spurned a cottage industry of YouTubers like and Taiki Maeda who’ve garnered a following by making the technically overwhelming approachable through shareable spreadsheets and click by click guides.
This is where Intents come in. “Intent based products are a move away from the transactional approach to DeFi,” Aperture’s McDavid Stoddard told
Intent based architectures enable the average DeFi user to gain access to “the types of execution and pricing that previously were only accessible to well capitalized prop shops operating with an in-house developer team. Even a devoted DeFi user armed with YouTube premium and tracking spreadsheets would fail to get the level of execution and expressibility that is possible with an Intent based DeFi UX.”
With Aperture’s Liquidity Intents solution, built atop Uniswap V3, the user declares an end goal for their LP position and the transaction is only executed if it matches the user’s desired outcome. A user can, for example, declare that they want their liquidity position to rebalance once ETH hits 3,500 USDC per ETH, rebalance to a 50:50 ratio at a range of 3,400 to 3,600 USDC, and specify the acceptable range of price impact from swaps and gas fees.
A decentralized network of solvers then works out possible transaction flows which are then simulated and ranked. Aperture has launched with one centralized solver, with third-party solvers including one from joining the network in Q2.
As well as building out its network of third-party solvers, Aperture’s future roadmap includes several more key components to a fully fledged Intent based DeFi UX. A Large Language Model (LLM) will enable users to express their intent in natural language and have it mirrored back to them in a highly readable Domain Specific Language (DSL) created by Aperture, which can then be delivered to solvers as the user’s “declaration of truth.”
Stoddard likens the UX to ordering a pizza. “You tell the operator, ‘I want whatever your biggest meatiest pizza is.’ The operator will mirror back to you, ‘Okay, you want our XL meat lovers’ pizza?’ And then the user says, ‘Yes, that one!’” Aperture’s UX will enable the user to declare their DeFi Intents in similar terms. “The user will say, ‘I want you to rebalance me to the highest yielding pool,’ and then the Intents chatbot will mirror back to them the highly readable DSL output which the user will be able to eye-ball and confirm.”
