PromptQL

PromptQL is a tool that helps AI models work with data more effectively. It creates query plans that can be adjusted during execution based on intermediate results or user instructions. This makes it better than traditional systems that cannot adapt their strategies on the fly.

Key Features

Dynamic Query Planning

PromptQL uses simple control structures like if-else statements to handle different situations. This makes it more flexible than traditional systems.

Seamless Data Integration

Hasura DDN connects to various data sources, including SQL databases, NoSQL stores, APIs, and SaaS platforms. This lets AI models access all data under one schema, making it easier to integrate different data sources.

Complex Computations and Data Transformations

PromptQL has a runtime environment that allows AI to perform complex calculations and data transformations. It handles large datasets by processing data in batches and can call another AI model to analyze text data.

End-to-End Task Completion

PromptQL enables AI to perform actions and update data by executing commands. This allows for complete task completion, unlike many AI applications that can only retrieve data.

Memory Artifacts for Large Datasets

To handle large datasets and keep conversation context over long interactions, PromptQL uses memory artifacts. This ensures consistency and accuracy without overloading the AI model.

Transparency and Explanations

PromptQL provides detailed execution plans and can explain each step. This transparency helps users verify the AI''s answers, ensuring trust and reliability.

Role-Based Access Control

Hasura DDN has fine-grained, role-based access control for data querying. This ensures that all data access follows predefined authorization policies, addressing security concerns.

Benefits

PromptQL offers a more dynamic, flexible, and transparent approach to data querying and task completion. It ensures consistency, accuracy, and security, making it a valuable tool for developers and data analysts.

Use Cases

Building Assistants

PromptQL can be used to build assistants that interact with data more effectively. For example, a GitHub issues assistant can manage and retrieve information from GitHub issues.

E-commerce Data Analysis

PromptQL can analyze e-commerce data by creating dynamic query plans that span multiple tables. It can identify top customers by revenue who have submitted reviews and filter results based on specific criteria.

Learning and Development

PromptQL acts as an expert, providing sample code and clear explanations for each step. This makes it a great tool for learning and development, helping users understand the processes behind the queries.