JetBrains uses OpenAI’s API to build its fastest-growing product ever.

Frame 1532576563
March 21, 2024
Jetbrains

JetBrains creates intelligent software development tools used by more than 15 million engineers around the world. Its product suite includes popular Integrated Development Environments (IDEs), such as PyCharm, IntelliJ IDEA, as well as the Kotlin programming language. 

Driven by its mission to remove tedious tasks from developers’ workflows, JetBrains incorporated OpenAI’s API into its AI Assistant product. The response has been striking—a notable 77% of developers reported feeling more productive, with 55% finding more time to focus on more engaging tasks.

Embedding AI into developer software

JetBrains IDEs have historically included features such as code completion and code inspection that make developers more productive. These features help developers through the lifecycle of software development, including writing code, testing code, and building projects in various environments. 

These tools were based on heuristic models that are smart and context-aware. Since JetBrains was already building tools that provide intelligent suggestions, exploring AI was a natural extension—especially given the recent advancements in generative AI, where features like autocomplete-style code completions and contextual code suggestions are transforming how developers work. 

JetBrains recognized the opportunity to take a significant leap forward and developed its AI Assistant, a new product that integrates with OpenAI’s API into Jetbrains’ IDE products.

Using OpenAI’s API to fast-track the software development process

JetBrains believed that LLMs could provide immense value to their customers if integrated into their IDEs, but wanted to explore their options before committing to a chosen provider. 

JetBrains considered all the leading LLM providers, but chose OpenAI as first LLM provider for the following reasons:

  • Advanced reasoning capabilities: JetBrains found OpenAI’s API to have the most intelligent features, with its advanced reasoning capabilities standing out in particular.
  • Technical customer support: Vladislav Tankov, Team Lead of JetBrains AI, emphasized the stellar technical support provided by OpenAI’s technical team. This guidance was pivotal when JetBrains encountered challenges during the integration process: “We worked with developers who really care about what we create and how we create it,” he said.  
  • Superior technical performance: In JetBrains’ comparative analysis of LLM Providers, OpenAI’s API stood out in terms of technical criteria such as, latency, accuracy, and throughput.

We worked with developers who really care about what we create and how we create it.

Vladislav Tankov, Team Lead of JetBrains AI

Having decided on OpenAI's API as their LLM provider, JetBrains was then able to unlock a range of capabilities in its AI Assistant product: 

  • Developers can offload mundane and less creative tasks, like writing tests and naming variables, to the AI Assistant.
  • Developers can write prompts in natural language and receive code suggestions when writing business logic.
  • The AI Assistant can generate documentation and write commit messages.
  • The AI Assistant can explain code and refactor code snippets.
  • The AI Assistant can clarify what errors mean and suggest fixes to address issues.
Jetbrains Ai 1280x720

JetBrains AI Assistant accelerates developer performance

JetBrains AI Assistant has become the fastest-growing product in the company’s 24-year history with outstanding levels of customer satisfaction. Developers who use the product highlight: 

  • Unprecedented productivity gains: “I’ve been a developer for 30 years. I can’t think of anything that has provided a bigger productivity boost than JetBrains AI Assistant,” one respondent said. 

  • Spending less time on information search: 78% of surveyed developers report spending less time searching for information, giving them more time to spend on more engaging tasks.

  • Reduced switching costs: “JetBrains AI Assistant operates inside the IDE, unlike other assistants that require an open browser,” one engineer said, which simplifies the developer workflow and makes it easier to complete tasks.

  • Increased efficiency: 77% of polled developers report feeling more productive with their time, while 71% report being able to complete tasks more quickly.

  • A brainstorming tool: One engineer emphasized how JetBrains AI Assistant has become a helpful tool for getting unstuck: “I love it. It helps with imposter syndrome because you can just ask the LLM for help.”

Building the next generation of context-aware AI coding tools

JetBrains is working on several new features to further accelerate software development: 

  • Project structure generation: The team has built prototypes that generate customized project directories from text prompts. Initial results look exciting.
  • Automation of workflows in the terminal: The team sees huge potential in automating various common terminal tasks that are necessary but mundane.
  • Tight integration with JetBrains code insight: “Our IDEs know a lot about the codebase of a user's project, so we are looking to provide this context and capability to LLM providers for completely new features,” said Vladislav Tankov, Team Lead of JetBrains AI.

Vlad Tankov Jetbrains Ai