|
| 1 | +Date: 27-06-2026 |
| 2 | + |
| 3 | +Duration: 1 hr 48 mins |
| 4 | + |
| 5 | +==== Topics Discussed |
| 6 | + |
| 7 | +* Janvi showcased a vibe-coded agent designed to help newcomers navigate production systems. |
| 8 | + ** She utilized link:https://www.langchain.com/[LangChain^], a link:https://www.trychroma.com/[Chroma^] vector database, and RAG to simplify the onboarding process and make complex codebases more understandable without needing to read the whole codebase. |
| 9 | +* Bhavesh shared his learning journey and a hilarious new side project. |
| 10 | + ** For learning AI fundamentals, he highly recommended checking out courses on Udemy, as well as Stanford and Harvard-based learning materials. |
| 11 | + ** He also demoed a joke side project: a to-do list for a to-do app, complete with a purposely frustrating ragebait login page. |
| 12 | +* Urmil (Gamiltron) discussed his backend learning journey. |
| 13 | + ** He shared a sign-in/sign-out system he built for a to-do app as a learning project. |
| 14 | + ** He asked the group for advice on whether his first major project should be complex and hardcoded or something smaller, which sparked a great group discussion on learning paths, effort, and career progression. |
| 15 | +* Alpesh shared his recent experiences from Draper Startup House in Bangalore. |
| 16 | + ** He highlighted his key learnings and the interesting founders he networked with during his stay. |
| 17 | +* Siddhesh discussed his ongoing project of building a custom Python interpreter. |
| 18 | + ** He shared and referenced specific books related to interpreter design and mechanics in Python. |
| 19 | +* Ayush gave an overview of his work at link:https://sherlocks.ai[Sherlocks.ai^]. |
| 20 | + ** He explained his role and answered questions from the group about whether the Sherlocks platform autonomously resolves incidents. |
| 21 | +* The group had a deep discussion around open-source ecosystems. |
| 22 | + ** They shared actionable tips on how to step up as a maintainer, efficiently manage projects, and make high-impact contributions. |
| 23 | +* The conversation then shifted to advanced AI development and tooling. |
| 24 | + ** The discussion covered the use of `agents.md`, context layers in mapping, and leveraging specific link:https://claude.ai/[Claude^] skills for building better agents. |
0 commit comments