Attended ChaiCode’s Wednesday webinar ☕️
A Claude Code AI masterclass learned where and how to use AI while building software.
Best part? Wednesday sessions are free to attend
@ChaiCodeHQ@Hiteshdotcom@piyushgarg_dev@nirudhuuu
@Hiteshdotcom@ChaiCodeHQ Thank you so much, sir! 🙌 Your comment means a lot.
Yes, I tried to make it as real-time as possible on Mac. Still lots to improve, and I’ll keep working on it whenever I get time.
Really grateful for the encouragement. ❤️
Inspired by @Hiteshdotcom , I built a speech-to-text app for my own daily use.
I couldn’t find an app that supported clear Gujarati transcription the way I wanted, and since I use Gujarati every day, I decided to build one myself.
The video has less Gujarati than I originally planned 😅, but Gujarati support is actually what led me to build this.
I’ll keep improving it as I get more time, but honestly, building something on the weekend and then actually using it every day… that’s peak engineering feeling. 🚀
Attended Wednesday’s webinar by @ChaiCodeHQ ,
“AI vs Engineers: Who Will Write the Code?”
What I took away:
• How interview processes are changing these days
• New rounds and approaches in interviews
• How engineers can instruct AI to ship better
• Ended with a fun quiz too!
@Hiteshdotcom
What a journey. ❤️
Started this cohort knowing very little about GenAI, and ended it learning, building, breaking, fixing, and creating things I never thought I could.
So many concepts, projects, late-night sessions, laughs, and memories along the way.
From Transformers and RAG to AI Agents and actually building projects. it has been one hell of a learning experience.
A huge thank you to @piyushgarg_dev , @Hiteshdotcom and @surajtwt_ for making every session so engaging, practical, and fun. And to everyone in the cohort who made this journey even more special. 🫶
Grateful for the learning, the people, and the memories.
@ChaiCodeHQ@nirudhuuu@BlazeisCoding
#GenAI
Today in class, I learned how to build a CLI AI agent using the Claude SDK. 🤖
I’ve started building my own CLI with features like plan mode and more coming soon.
For now, I’ve mainly worked on the UI — the real agent functionality is the next part. 🚀
@ChaiCodeHQ@surajtwt_
Learned how to build an AI Pitch Deck today.
Now it's time to put that learning into practice using Inngest to turn user ideas into carousel slides and ImageKit to store generated images.
From learning the concepts to building the project. 🚀
@ChaiCodeHQ@surajtwt_
Today I got a little deeper into voice agents 🎙���
Learned how STT, TTS, LLMs and streaming work together, plus handling interruptions and outputs.
Really interesting to see how these pieces come together to make AI conversations feel natural and real-time.
@piyushgarg_dev
Learn an Agent Skills today.
- Build one yourself.
- Learn how to publish it.
- Then install and use it with your AI agent.
Learn → Build → Publish → Install → Use.
@piyushgarg_dev@Hiteshdotcom@ChaiCodeHQ
Today we explored MCP (Model Context Protocol) and how AI agents interact with external tools. 🚀
Configured an MCP server locally using stdio and also connected one over the network.
A great session to understand how agents can actually use tools.
@ChaiCodeHQ@piyushgarg_dev
Started building Chaibook LLM—our take on NotebookLM.
Kicked off the backend, improved authentication, and laid the foundation for what's coming next.
Suraj sir showed up in a Spider-Man suit. Peak developer energy. 🕷️😂
@ChaiCodeHQ@surajtwt_@Hiteshdotcom
Today's topic: Memory in AI Agents.
Learned how agents manage short & long-term memory, decide what to retain or forget, and use vector stores & graph databases to build lasting context.
Memory makes AI agents feel personal.
@ChaiCodeHQ@piyushgarg_dev@Hiteshdotcom
Today's rabbit hole: Vectorless RAG.
Learned how modern AI systems can retrieve knowledge directly from PDFs, Drive, local folders, and web sources—without relying solely on vector databases.
RAG is evolving faster than I expected.
@ChaiCodeHQ@piyushgarg_dev@Hiteshdotcom
Today's session went beyond basic RAG.
Learned advanced RAG pipelines—query rewriting, HyDE, query routing, guardrails, ranking, and choosing the right data source before the LLM responds.
@ChaiCodeHQ@Hiteshdotcom@piyushgarg_dev