In an exclusive interaction, Karan M.V., Director, International Developer Relations at GitHub, shares insights with OSFY’s Saba Aafreen on how AI and open source are reshaping software development, highlighting the rise of agentic AI and its impact on how software is built, along with India’s growing role in the global developer ecosystem.
Q. Could you briefly share your journey at GitHub and how you have seen the platform evolve over the years?
A. I have been with GitHub for nearly six years, joining at a time when it was primarily known for version control, code hosting, and open source collaboration. Since then, the platform and the developer ecosystem, especially in India, have expanded significantly, with the number of developers on GitHub growing from around 4 million in 2020 to over 27 million today. Over the years, GitHub has evolved into a much broader software development platform that combines automation, collaboration, and AI driven capabilities, including agentic AI, enabling enterprises, teams, and individual developers to build software more efficiently. Being part of this transformation has given me the opportunity to closely engage with developers, customers, and open source communities worldwide while witnessing how AI is reshaping the future of software development.
Q. What key shift in developer behaviour drove this change?
A. The biggest shift has been the rise of open source, which encouraged developers to collaborate, contribute, and build across software, documentation, education, and AI projects. India now ranks second globally in open source AI contributions. At the same time, AI and agentic AI have made software creation accessible beyond traditional coders, enabling people from non-software backgrounds to solve real-world problems using tools like GitHub Copilot. This has democratised software development, accelerated collaboration, and significantly increased innovation and problem-solving through technology.
Q. How has AI changed the understanding of developer productivity?
A. Agentic AI is shifting developers from being coders to problem solvers and orchestrators of intelligent agents. Instead of manually writing every layer of software, developers now guide specialised AI agents across frontend, backend, databases, and testing while focusing more on creativity, judgment, accessibility, and real world impact. This has significantly improved productivity by enabling faster development, quicker problem solving, and higher quality software delivery. What we see is that this shift is also helping collaborative initiatives like the Open Healthcare Network scale open source healthcare solutions across hundreds of hospitals in India.
Q. Where does GitHub stand today in the AI driven software development stack?
A. GitHub sees AI, agentic AI, and security as core parts of the developer workflow rather than standalone add-ons. Tools like Copilot and AI agents are integrated directly into workflows such as issue creation, pull requests, code reviews, CI pipelines, and debugging. GitHub is also enabling new AI native workflows through tools like Copilot CLI and the Copilot app, allowing developers to work more efficiently across terminals, web platforms, and orchestrated AI agent environments while maintaining flexibility and choice.
Q. What are the biggest structural changes in GitHub’s developer ecosystem today?
A. One of the biggest changes at GitHub has been the shift from a version control platform to a complete AI powered software development ecosystem. GitHub expanded from code collaboration into automation through GitHub Actions, and now supports AI driven workflows across planning, building, testing, deployment, and operations. A major focus has been giving developers flexibility through model choice, agent choice, and workflow choice, allowing them to use AI models from OpenAI, Anthropic, Google, and others alongside tools like Copilot and Codex. Developers can now work seamlessly across web, mobile, CLI, VS Code, and other IDEs, making AI agents a core part of modern collaborative software development.
Q. How is debugging, architecture, and system design evolving in an AI assisted development environment?
A. Debugging, architecture, and system design are becoming more intelligent and collaborative in an AI assisted environment, where developer fundamentals remain critical for identifying what can go wrong and how to resolve it. At the same time, effective use of agentic AI allows developers to prevent and resolve issues more efficiently by clearly communicating system constraints and letting agents help design and build within those boundaries. With tools like Copilot integrated into platforms such as GitHub, agents can access repositories, observability systems, and error logs to correlate issues, trace root causes, and even propose fixes through pull requests. These changes enable a workflow where multiple specialised agents assist in debugging, code review, and architectural validation, while developers apply judgment to ensure system-level correctness and scalability.
Q. What differentiates successful open source projects from the thousands that never scale globally?
A. Successful open source projects are defined less by raw scale and more by how effectively they solve a specific problem for their intended users, whether niche or large scale. Key differentiators include clear problem fit, active maintenance that keeps the project updated for new features, performance needs, and evolving requirements, and a strong community that contributes, gives feedback, and helps others adopt and improve the project. Equally important is collaboration across developers, maintainers, organisations, and even other open source projects, since the ethos of openness and shared contribution drives long term success. Practices and guidance like those in open source guides also help maintainers improve project health and adoption.
Q. Are there any skill gaps despite India’s massive developer base?
A. The focus is less on skill gaps and more on continuous upskilling as the industry evolves. A key requirement today is the effective use of AI and agentic AI tools, which are becoming central to developer productivity. At the same time, strong fundamentals in computer science, software engineering, and system design remain critical for building scalable and reliable systems. As applications increasingly need to work across low bandwidth environments, diverse user needs, and accessibility constraints, problem solving and architectural thinking are becoming even more important. Overall, continuous hands-on learning, experimentation, and adaptation are essential to stay relevant in an AI-driven development landscape.
Q. Are there opportunities for student developers and universities?
A. It is one of the best times to be a student developer because the demand for software is increasing rapidly, which creates many real problems to solve. Students today have access to a full ecosystem of tools, from collaboration platforms like GitHub, to open source communities, to free developer resources such as the GitHub Student Developer Pack and AI tools like Copilot. These tools are not only helping students build software faster but also learn new languages, understand codebases, and explore system design in practical ways. Combined with strong community support and open access to infrastructure, students and universities now have everything needed to learn, experiment, and build real world solutions at an early stage.
Q. How are enterprises responding to AI assisted development, and what governance concerns remain?
A. Enterprises are largely focused on the productivity gains from generative AI and agentic AI, using them to accelerate product development, improve service delivery, and bring solutions to market faster. At the same time, their primary focus is on governance, guardrails, and responsible usage of AI within organisational workflows to ensure safety, compliance, and control over how AI generated code is used. Platforms like GitHub address this through tools like Copilot, which provide enterprises with built-in controls and governance mechanisms. This balance of productivity and oversight is reflected in adoption by major organisations such as Infosys, Wipro, TVS Motors, Air India, and Paytm, along with global system integrators establishing dedicated centres of excellence to scale AI driven development effectively.
Q. Are enterprises becoming more dependent on open source ecosystems than proprietary software?
A. Yes, enterprises across startups and large organisations are increasingly recognising the value of not just adopting open source but also actively contributing back to it. Many companies now have dedicated internal teams focused on how they use open source responsibly, including contributing to projects, funding initiatives, and sponsoring ecosystem development. Programmes like the GitHub Secure Open Source Fund bring enterprises together to pool funding and improve the security and sustainability of open source projects, with participation from organisations like Zerodha. This shift reflects a broader trend where enterprises are not only consumers of open source but also active contributors, helping strengthen projects across fintech, healthcare, and software infrastructure ecosystems globally.
Q. Beyond AI platforms, what is driving this rapid growth in developers?
A. According to GitHub, a major driver is the idea that “anybody can develop,” where people no longer need formal software engineering backgrounds to build applications and solve problems. AI tools like Copilot have lowered the barrier to software creation, enabling individuals, teams, and organisations to rapidly turn ideas into working solutions. This democratisation of development, especially in a large innovation ecosystem like India, is accelerating experimentation, hackathons, open collaboration, and real world problem solving at a much larger scale.
Q. Are Tier 2 and Tier 3 cities becoming significant contributors to GitHub’s growth, and what trends are you seeing from India’s developers?
A. GitHub does not track developer activity at a Tier 1, Tier 2, or Tier 3 city level, but at a country level India’s developer ecosystem is showing strong and rapid growth. A key trend driving this is how AI tools like Copilot are lowering barriers for developers in smaller towns and remote regions, enabling them to build software even with limited language fluency or traditional engineering exposure. Open source collaboration is also decentralising participation, allowing developers from anywhere to contribute, fix bugs, and build projects without needing to be in major tech hubs. Together, AI and open source are making software development more accessible, inclusive, and geographically distributed, unlocking talent from across the country.
Q. Do you see India moving from a contributor economy to a product building economy in software?
A. GitHub believes India is already making that transition, with developers building globally adopted products, platforms, and open source projects rather than only contributing to existing ecosystems. Examples include HyperSwitch, Bruno, ERPNext, and the Open Healthcare Network, which has been recognised by the United Nations as a digital public good. The growing global collaboration around Indian projects, along with increasing funding, monetisation, and community support, reflects how India is emerging as a strong product and innovation driven software ecosystem.
