Developer advocates Paul Everitt, Dr. Jodie Burchell and Khalid Abuhakmeh take us through demonstrations of seven key features: refactoring and explaining code, automatic code completion, inline code generation from prompts, writing documentation, creating commit messages, and the integrated AI Chat, which acts as an LLM-based rubber duck right in the IDE.

Svetlana Zemlyanskaya, AI Assistant Team Lead, joins to explain how the AI Assistant works under the hood. She explains how this new product is able to create deeply context-aware prompts in order to get targeted answers from the underlying models. Vladislav Tankov, AI Platform Team lead, further dives into the technology behind AI Assistant, JetBrains AI service. He gives us an overview of how the AI service allows the AI Assistant to use a variety of large language models in a transparent and secure manner.

Finally, we’re joined by Max Shafirov. Max shares how the AI Assistant continues a long tradition of intelligence inside JetBrains tools, with an anecdote about how early code inspections were done in IntelliJ IDEA back in its early days.

Integration and Functionality

• Refactoring: AI Assistant helps in code refactoring by suggesting changes and explaining why they are necessary.
• Context Awareness: The AI Assistant deeply integrates into the JetBrains IDE to understand the code context, making better suggestions.
• Availability: The AI Assistant is available for all JetBrains commercial IDEs, including ReSharper and Fleet, through a subscription plan.

AI Chat

• AI Chat acts as a "context-aware rubber duck," providing answers tailored to the project’s context, making it more accurate than generic tools like ChatGPT.
• Users can prompt for explanations or code snippets, enhancing productivity.

Code Completion

• Demonstrated in WebStorm, the AI Assistant offers advanced code completion by analyzing surrounding code for more accurate suggestions.

Documentation and Commit Messages

• The AI Assistant can generate detailed and formatted documentation for code.
• It can also create commit messages by summarizing code changes, streamlining the process.

AI Service Architecture

• Vladislav Tankov explains that the AI Service orchestrates interactions between the IDE and multiple language model providers, ensuring the best model or approach is used. It also addresses security and data privacy, with plans to cater to enterprise needs.

Future and Enterprise Solutions

• JetBrains is developing solutions for enterprise customers, ensuring on-premises deployment and integration with existing cloud providers.