# SyncEcho: Intent AI Inference

With SyncEcho, Users can actually chat with the blockchain/dApps as well. It is powered by an advanced intent-centric architecture that prioritizes user control and flexibility while interacting with blockchains. Intent-based architectures, as the name suggests, focus on the user intent or desired outcome. Fundamentally, an intent in the blockchain represents a specific objective a user aims to accomplish. With intent, the user states what they want to do and leaves out how it should be done.

This is unlike the current systems in Web3, where users provide detailed instructions for each transaction step, making it a complicated and time-consuming endeavor. Consequently, users face constant frustration in achieving their goals and are vulnerable to exploitation by sophisticated hackers, bad actors, and scamsters.

Intents change this fragmented user experience. Users express what they want to achieve, leaving the execution intricacies to be managed by the blockchain protocol. 

> A key use case of this system is the introduction of **smart buttons**. When users receive notifications via the Sync Mobile app, **intent agents** present them with intelligent action buttons, allowing them to seamlessly take actions based on those notifications, such as approving a trade or sending funds, all without leaving the app.

## **AI-Powered Solver Network** 

SyncAI's solver network is driven by LLM inferences linked to the blockchain to discover counterparties and efficiently execute intents:

1. **Solver APIs**: The network incorporates various specialized algorithms, each optimized for tasks such as analytics, swaps, trading, lending, etc.
2. **Intent Matching with LLM**: A fine-tuned LLM model is used to identify inputs that can be triggered by SyncAI's solver APIs. The matching process can be represented by:

$$
M(I)=LLM(I∣θ,A)
$$

where *M(I)* denotes the matching algorithm's result for the intent (I). *LLM* represents the fine-tuned language model, conditioned on the model's parameters *θ* and the set of available blockchain APIs (A). This provides highly efficient and accurate intent identification, mapping intents to matched solutions.

**SyncAI's Intent Types:**

* **DeFi Intents:** SyncAI allows users to specify intents for decentralized finance (DeFi) operations like trading, lending, and liquidity provision. These intents simplify complex DeFi interactions by predicting the user's desired outcome while automating counterparty discovery and settlement.

{% content-ref url="/pages/EiP2iiYDjDQGULdxOZfj" %}
[DeFi Intents](/features/syncecho-intent-ai-inference/defi-intents.md)
{% endcontent-ref %}

* **Blockchain Explorer:** Intents allow users to interact with the different blockchains to explore network data, track transactions, and monitor on-chain activity in natural language. These intents streamline blockchain exploration, providing a user-friendly way to access comprehensive on-chain information.

{% content-ref url="/pages/YB8TyUEYpKK90058avDo" %}
[Cardano Explorer](/features/syncecho-intent-ai-inference/cardano-explorer.md)
{% endcontent-ref %}

* **Governance:** SyncAI enables an intelligent way to interact with Governance DAOs, dReps, Project Catalyst, and more. Users can create intents to submit or explore proposals, vote on projects, and engage in discussions, enhancing their participation in the decision-making process of the Cardano ecosystem.

{% content-ref url="/pages/x7Hpdi9RxYAjYHCa3gE7" %}
[Governance](/features/syncecho-intent-ai-inference/governance.md)
{% endcontent-ref %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.syncai.network/features/syncecho-intent-ai-inference.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.