Tabnine, an AI code completion assistant, now provides developers with suggestions for long snippets and targeted line code completions directly within Visual Studio Code or IntelliJ Idea IDEs.
In addition, Tabnine announced today that it has raised an additional $15.5 million in funding from investors including Qualcomm, Samsung Next Ventures and TPY Capital, bringing the company’s total funding to date to $32 million. The infusion of funds will be used later this year to add support for additional AI models and programming languages, the company said.
Autofill Solutions such as Tabnine reuse an organization’s common coding patterns to reduce unnecessary toil from developers, said Jason English, an analyst at Intellyx. Applied AI use cases are already present in the software industry for accelerating high-speed data enrichment, code refactoring and process automation, so it’s no surprise to see AI permeating the developer’s IDE, he said.
AI is the future of development
With the sheer number of patterns and iterations in software development, Tabnine believes AI is the future of development, said Dror Weiss, CEO of Tabnine. One reason for the success of AI-powered assistants is that a developer’s life is more difficult these days, he said, as developers must tackle limitless tools rather than living in the universe of one specific language.
An AI assistant puts code writing on track, suggests best practices for writing code, and potentially prevents a developer from writing code that doesn’t follow best practices, Weiss said. By adding more precise AI capabilities that target specific languages, developers can explore a new world of possibilities, he said.
In addition to providing a guiding hand for writing code, AI allows a developer to eliminate mundane tasks and focus on more meaningful tasks like analysis, Weiss said, making a developer’s work more meaningful and interesting.
Tabnin stands out from the crowd
Tabnine’s competitors previously included Israeli startup Codota, which took over competitor Tabnine end of 2019 and combined the two similar models into one product under the Tabnine name.
While Kite Team Server runs on a company’s GPU-equipped servers instead of a laptop’s CPU, Tabnine’s revamp means its versatile AI can run in any environment, including on an individual developer’s machine or in the cloud. Weiss said.
Another thing that sets Tabnine apart is its goal to democratize AI model training and make it available to everyone, Weiss said. Developers aren’t limited to Tabnine’s AI models, he said, because anyone can train their own models by connecting to GitLab, GitHub, or Bitbucket and then training the AI model that reflects best practices for a specific project.
By allowing developers to train their own AI, Tabnine is one step ahead of their competitors, said Diego Lo Giudice, vice president and chief analyst at Forrester Research. “This allows development leaders to scale up good internal coding practices and have more control over where the source code comes from,” said Lo Giudice, “especially when it’s raised as a concern by their executives.”