Meet Jules, Google's AI coding agent that could change software development workflows

Indian Express

5 days ago

As AI shifts from a niche concept to a common tool within the industry, it is also being used in software development. Software engineers and developers are increasingly using AI to write code. In fact, OpenAI's ChatGPT has become a widely adopted AI-powered tool among developers, with 82 per cent reporting regular use, according to data from Statista. GitHub Copilot is the second most popular option, used by 44 per cent of developers, while Google Gemini holds the third spot at 22 per cent.
But as AI becomes embedded in development workflows, it is giving rise to AI agents, tools that can perform tasks autonomously. From writing entire applications and debugging errors to generating documentation and automating CI/CD workflows, these intelligent assistants now support nearly every part of the software development and developers are fast adopting.
This is where Google's Jules comes in, an AI agent that reads your code, understands your intent, and gets to work. It is integrated with GitHub. Google announced Jules as a Google Labs project in December and introduced it to beta testers through a public preview at its I/O developer conference. Now, Jules is out of beta.
Here's everything you need to know about Jules.
Jules is essentially an AI coding agent, a digital helper that can perform tasks on your computer and is being positioned as the next big thing in artificial intelligence. It uses Gemini 2.5 Pro and allows developers to delegate tasks asynchronously via GitHub-integrated Google Cloud virtual machines.
You can think of it as an autonomous AI developer assistant that clones your GitHub repo into a secure Google Cloud VM, understands the full project structure, fixes bugs, writes tests, adds features, and operates asynchronously – basically you assign tasks, and it gets to work while you focus on other things. Unlike other coding assistants, Jules doesn't just suggest code: it plans, edits, presents diffs with full reasoning, and offers audio changelogs.
Not only does Jules work directly inside your repo, offering deep understanding of multi-file systems and real project logic but it also lets you assign tasks and walk away. Jules handles them in the cloud and sends back a reviewable pull request.
Powered by Google's Gemini 2.5 Pro, the AI agent demonstrates high-level reasoning, ideal for complex feature implementation and refactoring. But Jules goes further: it can summarise changes in spoken form, allowing teams to quickly catch up during standups or asynchronous reviews. Its native support for GitHub PRs and branching makes adoption seamless for any development team.
For developers working on frontend UIs, backend APIs, AI models, and end-to-end systems, the process can be time-consuming. This is where Jules comes in handy – automating sections of code that are repeatedly used across a program, writing unit tests, and suggesting modern dependency updates.
While tools like GitHub Copilot act as coding assistants and some agents can even handle full workflows like building apps or writing test cases without manual input: Jules goes a step further, becoming a true co-developer: not a human, but an intelligent AI agent.
Google updated Jules' privacy policy to clarify that data from public repositories may be used for AI training, while data from private repositories is not used. During beta testing, Google said thousands of developers used Jules, resulting in over 140,000 public code improvements.
Jules is aimed at AI enthusiasts and professional developers.
Jules is now out of beta, Google recently announced. The AI coding agent offers a free plan with 15 daily tasks. Since the public beta, Jules has logged 2.3 million visits worldwide, with India, the US, and Vietnam as the top traffic sources.