DeepSource: A revolutionary software that automates code reviews

If you ever talk to a developer and ask them their source of worry, they will take seconds to tell you it is the ‘development and implementation of a code without errors.’ This is exactly the problem that Jai & Sanket (Co-founders of DeepSource) wanted to address. With the help of static analysis, DeepSource can identify and fix bug risks, anti-patterns, performance issues, and security flaws on every commit and pull request.

Founded in December 2018, this uber-cool company has come a long way wherein it has analyzed more than 500K commits, resolved almost 10 million issues with an average analysis time of only 14 seconds. To begin with, how did this idea of pulling DeepSource come into existence? It is a story to tell.

Sanket & Jai have been involved in building software systems and managing engineering teams for over 6 years now. Out of their own struggle with coding, DeepSource was invented. DeepSource helps you find and fix issues during code reviews. DeepSource finds issues like bug risks, anti-patterns, security vulnerabilities, performance problems, style violations, documentation gaps, and deficiency of test coverage in source code.

DeepSource offers some very unique propositions:

• Less than 5% false-positives: All the issues reported by DeepSource are validated against the code's context -- removing false positives with great accuracy. Less false positives, along with high relevance of issues, is one of their major product goals.

• Reduce time spent on manual code reviews: DeepSource allows seamless integration with any team's existing workflows like pull requests and continuous integration helps the developers act on issues and fix them before a reviewer looks into them.

• DeepSource Autofix: DeepSource Autofix suggests fixes for issues detected and creates pull requests with the recommended changes. This helps one avoid the grunt work of fixing these issues manually.

• Secure, by design: DeepSource follows a comprehensive set of practices and policies to make sure the systems are secure.

However, developing something as unique as this came with its own set of roadblocks and challenges. Since only a very few companies in the world work with this technology, and probably no other company from India does this, hiring the initial folks who would shape the product was a major stumbling block.

DeepSource can be used by individual developers as well as engineering teams in companies. Any company that builds and maintains the software, and employs developers can avail DeepSource to maximize their efficiency and save their time.

This is not where they stop, DeepSource supports analysis on a few languages - Python, Go, Ruby, Java, and JavaScript, however, they are adding support for more programming languages in upcoming quarters. They are also adding integrations with source code hosting providers like Azure DevOps, Google Cloud Source Repositories, AWS CodeCommit, etc.

To sum it up for you, DeepSource is an automated code review tool that helps developers find and automatically fix bugs in code. It continuously analyzes source code changes and identifies bugs categorized as security, performance, anti-patterns, and potential bug-risks, and even formats code to follow style guidelines. DeepSource integrates with your existing code review workflow in GitHub, GitLab, and Bitbucket and runs analysis on every commit and pull request, automatically. DeepSource is free to use for open-source projects and they charge a subscription fee for private projects and on-premise Enterprise deployments.

To know more, refer to their website here: https://deepsource.io/