{"id":113516,"date":"2026-05-11T08:42:50","date_gmt":"2026-05-11T08:42:50","guid":{"rendered":"https:\/\/wp.krmangalam.edu.in\/blog\/?p=113516"},"modified":"2026-06-19T11:51:31","modified_gmt":"2026-06-19T11:51:31","slug":"gsoc","status":"publish","type":"post","link":"https:\/\/wp.krmangalam.edu.in\/blog\/gsoc\/","title":{"rendered":"GSoC 2026: KRMU B.Tech. Student Wins \u20b93L Google Stipend"},"content":{"rendered":"

Most engineering students spend four years preparing for placements. Some spend those same years building projects for the global open-source community instead. This year, that difference translated into a \u20b93,00,000 stipend paid directly by Google for the prestigious Google Summer of Code (GSoC 2026) to a third-year B.Tech CSE<\/a> student from Gurugram.<\/span><\/p>\n

Jayant Parashar, currently in his 6th semester of B.Tech CSE at K.R. Mangalam University (KRMU), was selected as a contributor to GSoC 2026. GSoC is one of the world\u2019s most competitive open-source programmes, where contributors are selected not through exams or campus recruitment drives, but through publicly visible coding contributions, technical proposals, and peer evaluation by experienced engineers.<\/span><\/p>\n

This blog is not simply about one student’s achievement. It is about what made it possible. A\u00a0 structured system that KRMU’s School of Engineering and Technology (SOET) has built for exactly this purpose, called the Open Source Mentorship Programme. But before we get into how this mentorship programme works, let\u2019s first shed some light on what GSoC actually is and why getting selected is as difficult as it sounds.<\/span><\/p>\n

What is Google Summer of Code?<\/h2>\n

Google Summer of Code is a programme that funds developers to contribute to real open source software projects. It is a stipend-based programme that Google runs for more than 12 weeks. GSoC 2026 is the 22nd season of this programme that has produced over 22,000 contributions from across 123 countries since its inception in 2005.\u00a0<\/span><\/p>\n

This year, 184 of the leading open source organisations participated in GSoC 2026. These participants publish a list of project ideas for which the contributors have to study the codebase, submit pull requests and then write a detailed technical proposal for the project they want to tackle. Mentors from the participating organisations review every proposal and interview shortlisted candidates before making their selections. Only the strongest proposals pass through the process.<\/span><\/p>\n

As a result, selected candidates earn a credential that carries genuine industry weight. A GSoC selection tells any hiring manager that this person can identify a real problem, design a technical solution, communicate it clearly to experts, and execute it independently.<\/span><\/p>\n

How Jayant Got Selected in GSoC 2026<\/h2>\n

Jayant Parashar, a 6th-semester B.Tech CSE student at KRMU’s School of Engineering and Technology, was selected by FOSSASIA, one of the world’s most active open-technology organisations. He was one of the contributors selected from 15,245 applicants from 131 countries in GSoC 2026.<\/span>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0\u00a0<\/span><\/p>\n

Jayant has a simple goal – making the lives of PyTorch developers easy. His GSoC 2026 project tackles something that quietly frustrates a lot of people working in deep learning, called manual logging. Every time you run an experiment, you’re tracking loss, accuracy, learning rates, etc., and it’s all done manually. This mundane approach slows you down, and it can be automated.\u00a0<\/span><\/p>\n

So he’s building native autologging directly into Visdom, a tool that many PyTorch developers already utilise for visualisation. Once it’s in place, the tracking happens automatically, in real time, without the researcher having to think about it. It is a clean, simple, and genuine time-saver.<\/span><\/p>\n

\n \n \"\"\n \n