Aperture Finance Decryption: An Initial Exploration of Intent Architecture

Written by: Aperture Finance

This is the first article in the Aperture Finance series on decoding the domain of intents. In this article, we will introduce intent narratives, the problems they can solve in the current DeFi space, and highlight the characteristics of Aperture as a leading project in intent architecture. This will be helpful for both newcomers to the intent domain and investors or practitioners who are already familiar with products like Anoma and have followed reports from Binance Research, Paradigm, and Delphi Digital.

Starting with Simple Scenarios

Imagine the future user experience of Aperture, where users can clearly specify the final state they wish to achieve from a transaction (and obtain the optimal solution through a bidding model):

• Hedge against 50% of the volatility risk of my ETH position, maintaining it for 30 days, using any means as long as the cost is locked within a 2% annual interest rate per ETH;
• Claim all eligible airdrop rewards for me, with a requirement that the intermediary fee does not exceed 1% (including related gas fees);
• Rebalance my liquidity positions across all chains and all DEXs (Uniswap, Maverick, and Sushiswap) to ensure that 60% of the funds are concentrated in the pool with the highest APR, 30% in the second highest, and 10% in the third highest.

In the current "transaction-centric" model, each of the above intents could give rise to a complete protocol, from financing to development, and then building its own community. However, developing a protocol for each case would take too long. Moreover, users are always constrained by the frontend and smart contracts. There is actually a better solution, but first, we must break out of the "transaction-centric" model.

Intent > Transaction

In the current DeFi model, people often use a familiar but very limited "transaction method" to achieve desired results, where users interact with dApp interfaces, trigger actions, and then await outcomes. This approach is largely similar to the dilemmas faced in healthcare systems—patients are faced with countless choices; they want to achieve specific medical goals but receive no guarantees of efficacy, and they cannot even control the final costs.

For example, Xiao Wang suffers from neck pain, and when seeking medical help, he has the following goals:

• Overall pain reduction of 40% to 100%;
• Costs below $500 (out-of-pocket);
• Relief within 3 months.

Similar to browsing on-chain solutions, Xiao Wang might directly look for surgical options, but the options he can think of will be limited by his level of understanding. He might encounter a surgeon claiming to alleviate pain through surgery, stating that the cost will not exceed $500. However, the results are often disappointing; the post-operative pain may not significantly decrease, but the actual expenses could far exceed $500.

In contrast, an "intent-centric" approach could revolutionize the entire process. In the same example, the patient could directly express their needs in natural language:

"Reduce neck pain by at least 40% within 3 months for under $500, prioritizing pain reduction, cost, and then time. Healthcare service providers must have a 99% trust rating. No matter the channel."

Assuming there is an intent-based healthcare platform, it would allow healthcare service providers ("solvers") to compete with each other to offer the best solutions. These providers could compete on a technical level, offering innovative therapies or more competitive quotes to stand out.

Ultimately, the best solution Xiao Wang receives might be a two-month treatment provided by a supplier with a 100% trust rating—daily sauna sessions plus bi-weekly rehabilitation treatments, resulting in a 90% reduction in pain for less than $200. (This article does not constitute medical advice; if you have neck pain, please seek medical attention promptly.)

Of course, this analogy has limitations in reality, as the medical field itself is inherently unpredictable. Costs and treatment times can be estimated, but the efficacy varies greatly among individuals. However, in our blockchain world, the situation is quite different: advanced process simulations, the application of zero-knowledge proofs (ZKP), and other technological advancements allow DeFi users to no longer be constrained by such uncertainties.

So, what conditions are necessary to make "intent-centric" a reality? Beyond mere talk, what functionalities and characteristics do intent solutions need to possess?

Core Components of Intent Architecture

From the end-user perspective, a truly intent-based user experience (Intents UX) has three key components that most DeFi dApps (execution layer) currently lack:

1. A brand new user interface that allows users to clearly express their trading goals. Users are no longer just faced with a bunch of options to choose "yes" or "no," but can explicitly describe the final state they wish to achieve.
2. A solver network that selects the optimal method for users through competition, flexibly achieving their goals.
3. Arbitration executors that can accurately rank the solutions provided by various solvers and execute the selected final solution. The solver network is responsible for selecting solutions based on rankings, ensuring solver accountability, and ultimately executing transactions between the dApp and the blockchain ("global composability").

Creating a brand new user interface requires a domain-specific language (DSL) that the frontend can use to convert intents into executable and verifiable intents on the blockchain. Establishing a solver network is also essential to provide users with better rates and results through competition.

In addition to these user-facing components, any intent architecture must handle the abstraction of a set of problems from users, such as:

1. Verifiability (global composability): The architecture needs a method to verify whether the solutions truly adhere to the user's intents. Standardizing the verification methods is the real innovation that releases "global composability." In other words, if there is a standardized way to verify the final state through arbitration executors, applicable across any chain or dApp, then users can have a truly "globally composable" intent network.
2. Ranking (zero-knowledge proofs): Beyond cross-chain and dApp verification, simulations and rankings of solutions are also needed. Some types of verification may be complex and cumbersome, making them unsuitable for on-chain proof. In such cases, zero-knowledge proofs (ZK) can reduce computational costs. These ZK solutions do not need to be built from scratch, as modular solutions like Axiom or Brevis can be leveraged.
3. Privacy (special permission solvers): Some intents may leak sensitive information that users wish to keep hidden (e.g., concerns about front-running attacks). A complete intent architecture can address this issue by enabling special permission solvers. These solvers are specifically designed to handle privacy-related intents, possessing special trust credentials that allow them to manage transactions requiring stricter security and confidentiality.
4. Deviant solver behavior (penalty mechanisms): To ensure that solvers fulfill transactions as expected and do not engage in any malicious behavior, the system needs to implement a penalty system. A key point of the intent architecture is its enforcement accountability system; if any malicious behavior occurs, such as transaction revocation or other malicious events, the intent architecture has the authority to impose penalties for such actions. This not only protects the integrity of transactions but also deters potential malicious behavior by penalizing solvers.

Aperture's Universal Intent Architecture

Returning to the previous example:

Rebalance my liquidity positions across all chains and all DEXs * (Uniswap, Maverick, and Sushiswap) to ensure that 60% of the funds are concentrated in the pool with the highest APR, 30% in the second highest, and 10% in the third highest.*

This example describes a universal intent applicable across all chains and DEXs. In the next article, we will elaborate on Aperture's grand vision in the intent space, which consists of a network empowered by a unified interaction interface and solver DAOs. In subsequent articles, we will continue to analyze Aperture's roadmap for 2024, detailing how we will gradually build a universal intent architecture from blueprint to realization. Stay tuned!