Sourcegraph automates ‘soul-crushing’ tasks with AI coding agents

0
Sourcegraph automates ‘soul-crushing’ tasks with AI coding agents

The software industry has long grappled with a paradox: the larger the development team and codebase, the slower the rate of progress. However, Sourcegraph believes its AI coding agents will overcome this bottleneck.

By automating repetitive and tedious coding tasks, Sourcegraph says its AI coding agents will unlock efficiency in enterprise software development while empowering humans to focus on higher-order work.

“This is industrialisation,” the company proclaimed, “breaking a complex process into 100 smaller tasks, then automating the soul-crushing parts and letting humans focus on the ones they’re good at.”  

Leave the boring work to AI coding agents

Sourcegraph’s new offerings include five AI coding agents starting with the “Code Review Agent,” which is now available via an Early Access Program (EAP) waitlist. This agent functions as a springboard for additional agents – Code Migration, Testing, Documentation, and Notify – that are expected to roll out in the coming months.  

Key capabilities include:

  • Code Review Agent: Already operational in large enterprises and available through EAP, it automates code review processes—offering instant feedback and quality assurance at scale.  
  • Agent API (EAP): A platform to help enterprises build their own custom agents atop Sourcegraph’s infrastructure, also accessible via waitlist.
  • Auto-edit in the editor: Goes beyond traditional autocomplete by suggesting edits across files, delivering instant code reviews, testing assistance, and documentation feedback.  
  • Unified developer interface: Seamlessly integrates code search, chat, and agents into the broader software development lifecycle tools like editors and code reviews.  

Unlike AI copilots that rely on predictive text or agents that claim to outright replace human developers and often miss the mark, Sourcegraph asserts its agents already function in production at some of the biggest enterprises:

  • Indeed: With over 700 developers, Indeed uses Sourcegraph agents to “catch major bugs, quality concerns, architectural issues, security vulnerabilities, performance problems, and more.

    According to Jeff Davis, VP of Engineering at Indeed: “Sourcegraph’s agents are a key part of our strategy in multiple stages of the SDLC [Software Development Life Cycle].”

    This automation is predicted to save hundreds of hours on manual reviews and debugging while potentially avoiding costly mistakes.

  • Booking.com: The world’s largest travel company, with 4,000 developers, is exploring an agent to migrate legacy code. A task anticipated to take over 10 years with human efforts could now shrink to months.

    “The agents we’re building with Sourcegraph are helping us reimagine software development,” said Bruno Passos, AI Innovation Lead. “At Booking.com, devs using Sourcegraph daily in the IDE are merging 30%+ more PRs [Pull Requests] every month than those who don’t.”

  • Priceline: Utilises Sourcegraph agents to triage bugs and alleviate the burden of tracking a vast platform’s moving parts. These agents leverage dispersed repositories of organisational knowledge, including deployment histories and Jira issues, to offer actionable insights to developers.  

The adoption of these agents offers not just tangible benefits, but also demonstrates a broader cultural shift in how companies approach software development amid increasing complexity.

‘Industrialisation’ is the future of coding

The term “industrialisation” is not merely a marketing buzzword for Sourcegraph—it underpins the company’s philosophy.

Rather than aiming to replace human developers entirely, Sourcegraph envisions humans and machines working together harmoniously, akin to the 250-year-long progression of man and machine in traditional manufacturing industries.

“Agents like this – that reliably automate specific tasks – are a transformative technology, and they work today in enterprise codebases,” the company explained. “History shows industrial progress comes from humans and machines working together, not replacing one with the other.”

Far from science fiction scenarios where humanoid robots replace developers outright, the AI agents aim to integrate into existing workflows and reshape the process of “getting work done” at enterprises. 

Automating repetitive duties such as testing, code reviews, and code migrations allows human developers to focus on creative problem-solving, innovation, and complex architectural decisions. Removing the drudgery from software development is likely to be welcomed by everyone.

(Image by Gerd Altmann)

See also: JetBrains Junie: An AI coding agent to enhance developer productivity

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Tags: agentic ai, AI, ai agents, artificial intelligence, coding, development, programming, sourcegraph

link

Leave a Reply

Your email address will not be published. Required fields are marked *