SQL Prompt 11

For Redgate AI-related questions that aren't answered here, please contact support@red-gate.com

Redgate AI is the collective term for all AI-powered features and capabilities that span across Redgate's software portfolio.

Individual Redgate AI features can have differing implementations, leveraging :-

Third party LLMs
Redgate's own hosted custom models
On-device machine learning solutions

Product	LLMs	2. Redgate models	3. On-device
SQL Prompt	✓	✓ (Query Index Analysis)
Redgate Monitor	✓*		✓ (Dynamic Alerting)
Redgate Flyway (coming soon)	✓*
Redgate Test Data Manager	✓		✓ (AI Data Classification)

*coming soon

Please consult the relevant product documentation for more details on the individual AI features.

Redgate AI features are opt in. End users will need to sign up and opt in to use Redgate AI features. Please consult the relavant product documentation on how to achieve this.

Please contact support@red-gate.com to add your domain names to a block list to prevent the use of the AI features.

Access to Redgate AI features can be enabled/disabled at a per-product level.

Redgate AI makes all AI requests to *.redgate-platform.com

If your organization enforces it, you willl need to set this subdomain in your firewall rules and allow lists to enable Redgate AI features.

No data stored in tables is sent.
Information sent to Redgate AI service is used only to construct the response and isn't retained by the service.
Some Redgate AI capabilities rely on context saved in a locally installed vector database.

Detail on the specifics of the information sent to the Redgate AI service is outlined in the per product documentation.

Redgate LLM features are powered by three different public AI models in combination to provide the best possible outcome (all subject to our privacy guidance). These are all hosted on Redgate’s Amazon Bedrock server.

Claude-3-7-Sonnet
Cohere.Embed-multilingual-v3
Amazon.nova-lite-v1:0

They are hosted in Frankfurt, London, Ireland, and Paris.

No, we do not use any external data
Our models are the same as those available for public usage (see models used)

The data is used to fulfill the needs of the Redgate AI-powered feature.
Redgate agrees to not use customers input data for training models other than for use by the customer. 
Redgate does not store customer data for longer than it needs to fulfill the AI-powered feature.

Further resources:

Resource	Description
Redgate EULAs	10.14 - use of Third Party AI in Redgate Software.
Product Specific Terms	8.3 Use of LLMs in SQL Prompt 26.4 Use of Third Party AI services in Redgate Test Data Manager
Redgate privacy notice	Legal information including our privacy policy, data protection and use of information.
Security at Redgate (pdf)	Document outlining Redgate’s security practices and processes.
Product security policy	Redgate's product security policy
Redgate Trust Center	Company information on privacy, security, certification, and compliance.
Information collected by Redgate products	Outline of information collected to improve product quality and features.
Third party AI providers	Subprocessors used by Redgate Environment services

Product specific information

SQL Prompt’s AI-powered features help you to write, review, and optimize SQL faster by combining its understanding of your database schema and SQL expertise, with the power of generative AI and machine learning. Write and modify SQL accurately using natural language, get clear explanations of unfamiliar code, and uncover performance-boosting index suggestions in seconds — all without leaving SSMS. The AI-powered features are opt-in to give you control over how and when you adopt AI into your workflow.

We have two AI-powered features in SQL Prompt: SQL Prompt AI and Query Index Analysis.

SQL Prompt AI uses public AI models and provides the following functionality:

Natural language AI prompts (Text-to-SQL) - Write and modify SQL using natural language. SQL Prompt AI transforms your words into schema-aware AI requests that will create a response from our Prompt AI services that will be directly inserted into your SSMS query window. Modifying SQL is easy too, just highlight your code and instruct SQL Prompt to change it using natural language in your query window. AI responses will work with SQL Prompt’s formatting, refactoring and code analysis features to ensure code is consistent and accurate.
Explain SQL - Make sense of legacy code, complex logic, or someone else’s work fast. With SQL Prompt’s Explain SQL feature you can highlight your chosen SQL and get an instant translation of what the code does in everyday language.

Intelligent context-aware suggestions - Prompt AI offers two types of contextual-aware suggestions to help you form your AI prompts:
- Initial personalized suggestions - If Generate initial suggestions using SQL History option is enabled, SQL Prompt AI will create context-aware suggestions based on randomly selected recent SQL queries from SQL History.
- Follow-up suggestions - These are automatically generated in the Prompt AI window after you first AI prompt is processed and provide contextual-aware potential continuations to your previous AI prompts (e.g. introduce additional filters on a SELECT statement).

Please see working with SQL Prompt AI for more details on existing functionality.

Query Index Analysis, when triggered by the user, leverages an internal Redgate LLM (and not a third party LLM), and offers the following functionality:

Improve SQL query performance - Get clear, actionable advice on what indexes you need to improve query performance. Query Index Analysis evaluates the potential impact of different indexes and delivers smart recommendations for improvement.

Query Index Analysis will provide a recommendation to the user, and does not itself change the database.

Please refer to "Which third party LLMs are used by Redgate AI".

The SQL Prompt AI feature Query Index Analysis also uses an internal Redgate ML model. The model that backs Query Index Analysis is trained on Redgate's proprietary internal datasets, which are derived from publicly available databases. No customer data is used to train this model.

Built into SQL Prompt: the industry-leading SQL code completion tool - For nearly 20 years, SQL Prompt has been a favorite among database professionals, thanks to its ability to make SQL coding faster, easier, and error-free from day one. The new AI-powered capabilities join a range of productivity-enhancing features — such as smart code completion, reusable code snippets, instant formatting, refactoring and code analysis — all grounded in SQL best practices and a deep understanding of your database schema.
- These new features are designed to work with other functionality within SQL Prompt to deliver accurate, context-aware results that speed up your coding. For example, before presenting AI-generated suggestions, SQL Prompt ensures they match your existing formatting style. Its built-in code analysis and auto-fix functionality also help resolve errors and ensure code follows best practices.

No extra cost or set-up - SQL Prompt’s AI-powered features are included in an active subscription to SQL Prompt, SQL Toolbelt and SQL Toolbelt Essentials (perpetual licenses are excluded). No extra costs or set up required. All you need to do is opt-in to use them.

Intelligent context-aware suggestions - As with any new feature, getting confident with using AI takes practice and trial and error. SQL Prompt’s initial personalized and follow-up suggestions (see working with SQL Prompt AI) give you a helping hand by analyzing your SQL and providing contextual, clickable prompts in the Prompt AI window to guide you on what to ask.

Query Index Analysis - Powered by Redgate’s machine learning model, Query Index Analysis evaluates the potential impact of different indexes and delivers smart recommendations for improvement. The result? Code that is optimized for speed from the start.

Underpinned by over 25 years of SQL Server expertise - The new AI-powered capabilities are backed by Redgate’s over 25 years' experience with SQL Server.

The functionality available in the Prompt+ Early Access Program is now part of the new AI features available in SQL Prompt.

SQL Prompt’s AI-powered features are available in version 11.1.0.18267 onwards with an active subscription to SQL Prompt, SQL Toolbelt Essentials or SQL Toolbelt (perpetual licenses are excluded). Once you have this version installed, you will be able to opt-in to the Prompt AI features by following instructions in enabling/disabling SQL Prompt AI.

To access the features, simply highlight some SQL in your query window and right click to show the SQL Prompt context menu. Then select any of the following: Open Prompt AI, Explain Selected SQL or Analyze Query Indices.

You can also open the Prompt AI window using the shortcut ALT+Z or the Prompt AI icon in the SQL query window if the option Show icon for editor selection is enabled. See working with SQL Prompt AI for more information.

Your machine needs access to the internet in order for this service to run.

We’d love to hear your thoughts on SQL Prompt’s AI-powered features. You can share any feedback by filling in our AI Features survey or by simply dropping us an email at sqlprompt-feedback@red-gate.com.

Keep an eye on our roadmap for a preview of what’s next for SQL Prompt. If you’ve got any requests, please add them to SQL Prompt’s User Voice.

Prompt AI has a fair usage rate limit in place of approximately 10 prompts every 5 minutes. Once this limit is reached, users will see a message notifying them that they’ve hit the cap.

If you find yourself reaching this limit frequently, we’d love to hear your feedback. Please email us at sqlprompt-feedback@red-gate.com.

Please check SQL Prompt AI data handling.

SQL Prompt 11

SQL Prompt AI FAQ

Product specific information