Video summary

GSoC 2027: Roadmap & Secrets | Watch Before Dec 2026

Main summary

Key takeaways

News and Commentary

Overview

This video is a Q&A-style discussion about preparing for Google Summer of Code (GSoC) 2027. It features a current GSoC mentor, who shares guidance from their own experience as both a contributor and a mentor.

Main Points and Advice From the Mentor

No “best time” to start GSoC preparation

The mentor emphasizes that there is no single ideal month to begin preparing. Instead, they highlight:

  • Consistency over time
  • Active communication with the community

They note that students can begin contributing as early as June for the 2027 cycle, and many start around October/November—both are acceptable. The determining factor is consistency.

What selection committees/mentors look for in PRs

Mentors judge far more than just the number of pull requests (PRs). They pay close attention to things like:

  • Whether the README and documentation match the code changes
  • Code cleanliness and maintainability (not just “it works”)
  • Avoiding poor practices such as duplicated logic
  • React-related concerns, including potential memory leaks
  • Evidence of system design thinking, including design patterns and scalability

How AI affects open-source contributions (and selection concerns)

The mentor observes that many proposals appear AI-generated during review. They also mention that attempts at plagiarism detection may be possible (e.g., via screenshot pasting), suggesting some contributors may act in bad faith.

Their view is:

  • AI can help with generating code and even supporting architecture work
  • However, engineering cannot be fully replaced by AI
  • Reviewers can often tell whether a contributor truly understands the work through:
    • PR quality
    • Context and communication around the changes

Debunking the claim that GSoC will end

A viewer mentions a rumor that GSoC might “die in 5–10 years.” The mentor responds that it’s not expected to be banished—Google/organizers are likely to adapt or find another way rather than cancel it outright.

Funding / Perks Overview (GSoC Stipend)

The mentor describes stipend amounts based on project size:

  • Small / medium / large project categories
  • They personally had a large project

For India, they mention:

  • $3,000 total
  • Split as:
    • 45% after midterm
    • 55% after final

For the USA, they estimate:

  • Approximately $6,000–$7,000 for a large project (as discussed in the session)

Personal Background and Journey

Education path

They explain that they completed a diploma (computer science) instead of going into class 12, partly because they didn’t enjoy maths and chose a non-medical track. Later, they entered BTech second year.

They also share their timeline of coding:

  • Started coding in school (around class 9)
  • Became heavily engaged during diploma (second year: coding almost every day)

GSoC journey

Their experience includes:

  • GSoC contributor in 2025
  • GSoC mentor starting from the next cycle
  • Contribution strategy through their company’s open-source work by selecting organizations/projects where they could genuinely help

Practical “How to Start” Guidance for Juniors (GSoC 2027)

1) Choose the right skills and tech stack

  • Learn the basic fundamentals of your target language/stack first.

2) Find suitable organizations

  • Use community resources (the mentor mentions a community-driven listing/filter approach, not necessarily official GSoC listings).
  • Choose 2–3 organizations
  • Join their communication channels (e.g., Zulip / Slack / Discord / Microsoft Teams depending on the org)
  • Review their maintenance standards

3) Don’t jump into building big features immediately

Instead:

  • Communicate first
  • Introduce yourself
  • Understand the codebase
  • Interact with maintainers

4) Focus on system design basics

  • Don’t rely only on coding
  • Practice system design thinking

5) Prioritize communication

Strong coding without communication can still reduce selection chances.

Common Mistakes the Mentor Would Avoid

A key mistake the mentor highlights is trying to build everything at once instead of targeting an MVP (minimum viable product) scope.

They give an example: social features like likes/comments shouldn’t be treated as MVP if the core product engine is located elsewhere.

AI-Oriented Closing Message

The mentor rejects the idea that AI means people don’t need engineers. Instead, they frame AI as an assistive tool—similar to earlier shifts in developer tooling—and stress the importance of upskilling to stay competitive.

Presenters / Contributors

  • Simjot (guest; GSoC mentor)
  • Host / interviewer (unidentified speaker: asks questions and moderates the session)

Original video