The addition of 2.2 million developers this year follows last year’s growth, where GitHub added 3.5 million developers, increasing India’s developer base to 13.2 million.The momentum is palpable, with the Indian developer base now having grown threefold from 5.8 million in 2021. “About 200,000 computer science graduates are coming out of engineering degrees, and that creates a convergence. India could be the leader, not only in software but also in AI. Children and adults alike will learn to code in their native language, leading to a prolonged groundswell of developers,” said GitHub CEO Thomas Dohmke.
During the December quarter of last year, developers in India owned over 36.9 million repositories on GitHub, compared to 27.4 million repositories during the corresponding period last year. Over 3.8 lakh developers from India made over 9.7 million contributions to open-source last year. In comparison, in 2022, over 3.2 lakh developers from India made over 8.1 million contributions to open source.
From reaping the benefits of AI pair programming to accelerating the next era of software development, GitHub is improving developers’ productivity. Last year, GitHub surpassed $1 billion in annual recurring revenue. Recently, Microsoft CEO Satya Nadella said GitHub revenue accelerated to over 40% year-over-year, driven by the growth and adoption of the GitHub AI developer tool Copilot. GitHub Copilot has 1.8 million paid subscribers and has been embraced by over 50,000 firms globally, including Autodesk, Dell, and Goldman Sachs. In India, GitHub counts Cognizant, Infosys, MakeMyTrip, and Paytm among its top customers. Among its recent AI features, Copilot Enterprise, tailors Copilot to a company’s codebase and allows developers to converse with it in natural language.
Dohmke grew up in Berlin in a suburb called Marzahn on the East side of the wall, before it fell. “The complexity of software systems has increased compared to the early ’90s. There wasn’t much software back then, but today you can’t imagine a world without software,” he said. When asked about how GitScore, along with GenAI, could be the ultimate litmus test in understanding developer proficiency, Dohmke said that in many interviews, instead of asking the developer to write code, interviewers ask them to use a copilot to write the same piece of code. “This shows their prompting and reasoning skills. It provides good insights into whether the developer not only understands how to work with models, but also how they’re validating the code that was written. This, in turn, allows developers to learn more about GitHub Copilot to upskill themselves in prompt engineering, learning how to use a large language model by writing prompts and refining them within the context of what they are working on. It’s a bit like learning how to drive a car,” he said. GitScore offers a comprehensive overview of a developer’s GitHub profile statistics and code analysis results.
Regulations like GDPR are upholding data privacy and have significant implications on how AI systems process personal data. “We’re actually interested in enabling open source and open research on GitHub, despite all these regulations (data privacy and AI). Because somebody with a great idea here in Bengaluru cannot afford the same compliance teams as we have at Microsoft or Infosys,” Dohmke added.
