AI Query

What is AI Query?

AI Query is a tool that converts natural language into SQL. Users first define the database schema, and then describe the desired data in plain language; the system generates the corresponding SQL code. The platform can also translate complex SQL queries into easier-to-understand explanations, and it offers suggestions for query optimization and indexing.

It is suitable for assisting with the writing of queries and learning SQL, but the statements generated should not be used in production databases without first being reviewed. In particular, statements related to updates, deletions, and structural changes must be tested in an isolated environment.

A one-sentence summary

AI Query generates SQL statements based on the database structure and natural language, and provides functions for explanation, optimization, history tracking, sharing, as well as a standalone API.

Core functions

• Natural language to SQL: Generate query statements based on English requirements.
• Schema awareness: Use table names, field names, and data types to select the correct objects.
• SQL interpretation: Translates complex queries into easy-to-understand natural language.
• Query optimization: Rewrite SQL to help the database optimizer come up with a better execution plan.
• Index recommendations: Analyze queries and identify potential indexes that may be needed.
• Database management: Save multiple database structures for selection during queries.
• Import table creation script: Use the Schema script to quickly establish the structure.
• Manual table creation: Add tables, fields, and types one by one on the interface.
• Query History: Saves the SQL statements that have been generated and interpreted in the past.
• Save and share: Organize frequently used queries and share the results.
• Developer API: Use RapidAPI to invoke functions for generation, interpretation, optimization, and indexing.

Which databases are supported on the web version?

The official introductory documentation lists four types of database systems that can be used via a web interface: PostgreSQL, MySQL, MariaDB, and SQL Server. The homepage of the official website also shows logos for additional databases and cloud services, but in practice it is necessary to follow the instructions provided in the console and the list of supported documents.

Tutorial on configuring database schema

1. Log in to the console and open the Databases section.
2. Select Add New Database and enter a recognizable name.
3. Choose between PostgreSQL, MySQL, MariaDB, or SQL Server.
4. Prioritize exporting from the database scripts that contain only the structure, without any business data.
5. Use Import From Script to paste the Schema and carry out the import.
6. Check whether the table names, field names, data types, and primary/foreign keys are complete.
7. After saving, use a simple query to verify whether the system understands the structure.

Tutorial on manually creating a Schema

1. Create a database entry and select the correct database type.
2. Click Create First Table and enter the table name.
3. Add field names and data types one by one.
4. Save the current table, then proceed to add other tables.
5. Add supplementary related fields and specify the relationships between tables in the prompt.
6. Test the generation of SQL using small-scale problems, and then correct any missing structures.

SQL generation tutorial

1. Go to Query Playground and select the database that has been configured.
2. Use one sentence to specify the fields to be returned, the filtering criteria, the time range, and the sorting order.
3. Add requirements for aggregation, grouping, deduplication, pagination, and handling of null values.
4. Click to generate and view the full SQL statement.
5. Verify the tables, fields, connection conditions, and database dialect being used.
6. First, execute it using a read-only account and test data, and check whether the results meet the expectations.
7. It is used in the official system only after executing the plan, using boundary samples, and conducting code reviews.

How to write accurate query requirements

• Return content: Specify which fields and calculation results are required.
• Data range: Specify date, status, region, or user group.
• Table relationships: Explain the associations such as orders belonging to customers, and details belonging to orders.
• Aggregation method: Indicates whether to use summation, average, count, or maximum value.
• Grouping dimension: Group by day, month, product, channel, or customer.
• Sorting and quantity: Specify ascending/descending order, top entries, and pagination.
• Handling of null values: Specify whether to ignore them, replace them, or count them separately.
• Database type: Ensure that the prompt is consistent with the selected SQL dialect.

SQL interpretation feature

The SQL to English Translator allows users to paste queries and obtain natural-language explanations, which is useful for reading legacy code and understanding connection and aggregation logic. These explanations can help identify potential issues, but they do not guarantee that the SQL code is correct or efficient.

• Understand the purpose of each table and the join conditions.
• Identify filtering, grouping, sorting, and aggregation logic.
• Explain to non-technical members what the query is doing.
• Cartesian products, missing conditions, or potential duplicates were detected.
• Prepare preliminary reading notes before the code review.

Query optimization feature

The optimization feature rewrites the SQL statements; it may break down complex queries, modify subqueries, or provide the database optimizer with a clearer execution path. The newly generated version must be tested to ensure that its results are equivalent to those of the original query.

• Compare the execution plans and the number of rows scanned before and after optimization.
• Verify the returned rows and aggregate results on the same data snapshot.
• Check whether the index actually exists and its maintenance cost is acceptable.
• Be vigilant against changes in NULL values, duplicate values, or join semantics after optimization.
• Conduct stress testing using data that is close to production scale.

How to use index suggestions

Indexes are recommended as a starting point for investigations; they should not be created directly in the production database. Indexes take up disk space, increase write costs, and may overlap with existing indexes.

• Check the selectivity of filtering, joining, and sorting columns.
• Compare existing single-column and composite indexes.
• Evaluate the write frequency and index maintenance costs.
• Verify the benefits using the actual execution plan of the database.
• Develop plans for creation, monitoring, and rollback.

Website price

The official website currently offers only the Pro package, available in monthly or annual billing options. The annual plan costs 100 dollars, which is 20 dollars less than paying monthly for a year; it also provides faster response times and priority support.

Common benefits of the package

• Unlimited generation of SQL queries.
• Unlimited SQL interpretation.
• Unlimited database schema.
• Save and share SQL without any limits.
• Unlimited queries to historical data are allowed.
• It can be canceled at any time.

How to choose between monthly and annual payment?

• Short-term learning or project testing: Choose a monthly payment plan first to reduce long-term commitments.
• If used continuously for over ten months, the annual total cost is lower.
• Pay attention to response speed and support priority: choose the annual payment option.
• Team trial: First, use a non-sensitive schema to verify accuracy and the workflow.
• Before making a purchase, check whether “unlimited” usage is subject to fair use policies and technical restrictions.

API features

• Generate SQL: Generate SQL based on natural language, schema, and database type.
• Explain SQL: Convert the input SQL into a description in natural language.
• Optimize SQL: Rewrite queries to improve performance.
• Suggest Indexes: Provides index suggestions based on SQL.
• Multiple model parameters: The document lists various context and model options.
• Multiple database dialects: The API supports eight different data source types.

How to use the API

The official API is made available through RapidAPI; to use it, a key generated by this platform as well as a Host request header are required. A Pro subscription for the web version does not automatically entail access to the API quota, and developers must check the plan details and restrictions on the RapidAPI subscription page.

1. Find the AI Query interface on RapidAPI and select the appropriate subscription plan.
2. Save the key; do not include it in the frontend code or in public repositories.
3. Select the endpoint for generating, explaining, optimizing, or suggesting indexes.
4. Enter the prompt, model, database type, and required schema.
5. Perform format validation, SQL parsing, and security checks on the responses.
6. Run it on a read-only test database, and record the calls and their execution results.
7. Design fallback mechanisms for timeouts, rate limiting, error responses, and model anomalies.

Which users are it suitable for

• Beginners in data analysis: Convert business problems into initial SQL drafts.
• Product and operations staff: Consult basic data with the assistance of engineers.
• Software developers: Quickly write and interpret queries they are not familiar with.
• Database administrator: Obtain clues for optimization and indexing.
• Teachers and students: Learning different SQL structures and dialects.
• SaaS development team: Adds controlled natural language query capabilities via API.

Typical use cases

• Generate queries for sales, users, orders, and retention analysis.
• Explain complex SQL in historical systems.
• Rewrite the PostgreSQL approach into a preliminary draft in other database dialects.
• Optimize slow queries and explore opportunities for indexing.
• Build an SQL generation interface for the internal data assistant.
• Prepare query prototypes for reports and dashboards.

Scenarios that are not suitable for direct use

• Unverified execution of delete, update, and structural change statements.
• Allowing unauthorized users to access sensitive production data.
• Upload the complete production data as prompt content.
• Rely on AI to determine financial, medical, and regulatory standards.
• Database operations that lack a testing library, as well as backup and rollback mechanisms.
• Core systems that require absolute accuracy in queries and stable performance.

Database security recommendations

• Prioritize importing only the Schema, without uploading actual business data rows.
• Use read-only, minimally privileged database accounts that can be revoked at any time.
• Remove keys, personal information, and trade secrets from field names and annotations.
• It is prohibited to use AI-generated statements to bypass code review.
• Set timeouts, limits on the amount of scanning, and limits on the number of result rows for queries.
• In the production environment, record the executor, time, statements, and scope of impact.
• Regularly replace API keys and database credentials.

Privacy and data deletion

The privacy policy states that the platform will store the registered email address, the database schema, as well as only the minimum amount of database information necessary to provide the services; it also keeps regular logs and cookies. If a user requests that their account be deleted, the platform commits to deleting the personal information, database credentials, and result caches within 4 working days, and to removing the relevant data from both production and backup systems.

The wording of certain parts of the policy document is not clear enough; before uploading sensitive schemas, companies should explicitly confirm the scope of storage, the handling by third-party models, the data areas involved, and the proof of deletion.

Product advantages

• Generating SQL by first combining with a Schema yields more accurate results compared to context-free querying.
• Generation, explanation, optimization, and indexing suggestions are all consolidated in one platform.
• It supports two ways to define a schema: script import and manual table creation.
• Monthly and annual subscriptions have simple structures, with prices that are easy to understand.
• Provides query history, saving, and sharing options.
• APIs can be used to develop controlled data applications.
• The official documentation includes introductory tutorials and request examples.

Usage restrictions

• AI may still choose the wrong fields, connection conditions, or aggregation methods.
• Generating SQL does not equate to performing performance and security checks.
• The web-based document explicitly supports only four types of databases.
• The API is provided through a third-party platform, and it requires a separate subscription as well as the management of keys.
• The claim of “unlimited usage” on the official website does not mean that there are no technical restrictions.
• It is necessary to upload or enter the Schema, as sensitive structures pose privacy risks.
• Query optimization and indexing suggestions must be verified using actual execution plans.
• There is no explicit guarantee on the official website regarding the performance of Chinese language prompts.

GitHub and the open-source status

AI Query has an official GitHub organization that has been verified by the website, but the only content made available publicly is the repository for organization configuration; there is no source code for the product. The SQL generation model, backend services, web applications, or SDK implementations are not available publicly.

Therefore, the AI Query platform itself is not open source; just because there is an official GitHub repository for it, it cannot be considered an open-source project.

Basic information

Recommendation score

Recommendation score: 4.2 / 5. It is suitable for users who wish to improve the accuracy of SQL queries through the use of schemas, and who need explanations, optimizations, and API support; however, ensuring the security of production databases and the correctness of the results still requires the attention of professionals.

Frequently Asked Questions

Is AI Query free?

The official website does not list any free plans; the Pro plan costs $10 per month or $100 per year.

Is it possible to use it without knowing SQL?

A draft can be generated, but the user still needs to learn basic SQL or have a professional review it.

Which databases are supported?

The web-based documentation explicitly supports PostgreSQL, MySQL, MariaDB, and SQL Server.

Can you explain the existing SQL?

Yes, the SQL to English feature will explain the query logic in natural language.

Can slow queries be optimized?

Optimized versions and index recommendations can be generated, but equivalence tests on the execution plans and results are required.

Will it connect to the production database?

The platform requires information on the database structure; it is recommended to use minimal read-only permissions and to avoid direct access to the sensitive production environment.

How long does it take for the data to be deleted after the account is removed?

The privacy policy states that relevant personal information and database data will be deleted within 4 working days upon receipt of a request.

Are APIs provided?

It is available; the API can be subscribed to and called through RapidAPI.

Is the API included in the Pro package?

The document does not specify any hierarchical relationships; it is necessary to check the prices and quotas separately on RapidAPI.

Is AI Query open source?

It is not open source; the official GitHub repository does not make the product’s source code available to the public.