Pratyush Sharma

Full stack Web and RAG Developer

[Pratyush Sharma @pratyush2374]

@emanuel_build Lo mismo aquí hermano

[Pratyush Sharma @pratyush2374]

@curious_vii yeah bro, same here

[Pratyush Sharma @pratyush2374]

Is the Apple Support only for namesake ?? I havent received any response from you @AppleSupport ...I tried contacting the authorized service centre, to my dismay they are charging Rs 31000 to change the display. Did I make a mistake by upgrading to iOS 26 ?

[Pratyush Sharma @pratyush2374]

I updated my iPhone 13 Pro Max to iOS 26, and now the screen has turned green. Please help fix this soon

@Apple @AppleSupport

[Pratyush Sharma @pratyush2374]

@AppleSupport @Apple
I updated my iPhone 13 Pro Max to iOS 26, and now the screen has turned green. Please help fix this soon

[Pratyush Sharma @pratyush2374]

@vikashsing45 @tim_cook @BlakSteveJobs @PS5OnlyUAE Hi, any update on this
Did they do it for free ??

[Pratyush Sharma @pratyush2374]

Let’s push the limits of what local AI assistants can do.
Would you want a deep dive on how I did this?
Drop a comment let’s build smarter AI together💬

#chaicode #AI #MCP #Claude #LocalAI #ContextManagement #TechInnovation #TripTuner #AIIntegration #SSE #SmartAI

[Pratyush Sharma @pratyush2374]

AI assistants aren’t smart. You just make them look smart.
That’s what I realised today after building my own MCP (Model Context Protocol) server.

I wired it up with Claude desktop, and suddenly, the assistant wasn’t just replying—it was replying with real, relevant context

[Pratyush Sharma @pratyush2374]

It feels like magic ✨
But it’s just clean context management
Standardized, structured, scalable
This has serious scope

🔹Now diving into SSE for live data 🔄

Try feeding your AI context.
Not prompt hacks. Not jailbreaks
Real, structured, useful context

[Pratyush Sharma @pratyush2374]

🌟 Built to deepen my understanding of RAG pipelines

👉 Try it out: https://t.co/rRSm8DnmO8
🔗 Source Code: https://t.co/RAg3zJOAqG

#FastAPI #React #GeminiAPI #Qdrant #ChaiCode

[Pratyush Sharma @pratyush2374]

🚀 Introducing ChAI – a simple yet powerful Retrieval-Augmented Generation (RAG) app, built from the GenAI cohort learnings by @piyushgarg_dev and @Hiteshdotcom .

Link: https://t.co/rRSm8DnmO8

Tap into docs smarter & faster. #RAG #AI #chaicode

[Pratyush Sharma @pratyush2374]

🤔 What is ChAI? It fetches relevant URLs & context from Chai Code’s documentation, then surfaces direct, meaningful answers

🔧 Tech Highlights:
• Frontend: Vite + React
• Backend: FastAPI
• LLM: Gemini API
• Vector store: Qdrant

[Pratyush Sharma @pratyush2374]

Every line of code made the theory click harder.
Nothing beats this learn-by-building energy 🔥

Really loving the hands-on pace of this GenAI cohort.

#GenAI #LLMs #RetrievalAugmentedGeneration #AIProjects #ChaiCode

[Pratyush Sharma @pratyush2374]

Quick build drop from my GenAI Cohort journey 💡

Day 6 with @piyushgarg_dev & @Hiteshdotcom focused on advanced RAG (Retrieval-Augmented Generation) techniques — and it seriously clicked.

Here’s what I built after 👇

[Pratyush Sharma @pratyush2374]

Techniques I experimented with:

🔹 Fan-out Retrieval
🔹 Reciprocal Rank Fusion (RRF)
🔹 HyDE (Hypothetical Document Embeddings)

Every method taught me something new.

[Pratyush Sharma @pratyush2374]

🚀 Ready to turn any dataset or PDF into an interactive AI assistant? Query transformation is your secret weapon

🗣️ Which technique will you try first? Drop a reply!

Interested? Grab the course here: https://t.co/L9Wz7xXXjQ

#GenAI #RAG #LLMs #VectorDB #AIFromScratch #ChaiCode

[Pratyush Sharma @pratyush2374]

🔄Day 6 in GenAI cohort with @piyushgarg_dev @Hiteshdotcom
Deep Dive into Query Transformation for RAG
Messy user prompts? Here’s how to turn noise into pinpoint queries so your AI retrieves exactly what you need

[Pratyush Sharma @pratyush2374]

5️⃣ HyDE (Hypothetical Doc Embeddings)

Ask your LLM to draft a “dummy doc” from the user prompt. Embed that text to enrich your vector search context and uncover deeper insights.