Purushottam Baghel

Everyone is building with Gen AI now. The hardest part is no longer building. It’s finding customers who actually care.

@getnachonacho 100%. Building is table stakes now. The real challenge is consistent outreach and finding where your audience is already actively asking for help. Showing up in those high-intent conversations early on is a massive cheat code for distribution.

Distribution is becoming the real advantage because building tools are easier to access now. Most founders should focus on audience, positioning, partnerships, and consistent outreach early. NachoNacho can also help founders explore discounted tools for marketing, CRM, and growth. Strong distribution turns a good product into a real business.

Finding customers is harder than building. Most people can build now. Very few know how to distribute.

After 5 months, I got back on X and started reaching out to people again. In just 15 hours: - 15 new followers - a bunch of great conversations - a strong reminder that building in public works This is the kind of momentum I needed. Back to showing up. Back to learning out loud.

Do you ever feel like your potential customers are already talking online… but you just cannot find those conversations? Finding people who actually need your product is hard. The real problem is not building.

@antonships Yup...I am working on runbyte.tech. what are you currently working on?

@PurushottamBagh Exactly, the deciding before the day starts removes the negotiation entirely. U wake up and just execute instead of spending the first hour figuring out what that first hour should be

The days you don't post aren't laziness. they're just days where the environment made starting harder than stopping Consistency is NOT ONLY a discipline thing, it's a setup thing. When the system is already there you just publish. When every day needs reinvention first you drift. I've felt both versions of that pretty clearly. What's a thing in your routine that makes showing up easier?

@_JohnBuilds_ I am building runbyte.tech. You can find your customer on reddit too

@PurushottamBagh A lot of great products fail quietly simply because nobody ever sees them 😭

Finding customers is harder than building. Most people can build now. Very few know how to distribute.

@kantariyaraj Looks Great..🎉🙌

Good morning builders, Working on building runbyte.tech (Reddit Relevance) a tool that helps founders find Reddit posts where people are already talking about problems their product solves. Give it a try. I’ll share what I’m building, what breaks, what works.

@AmazonHelp @amazonIN @amazon Same bot replying here and there in customer service chat. pathetic!!!

@PurushottamBagh Apologies for any unpleasant experience you've had with the delivery of your order. We'd like to escalate this issue to the concerned team. Please reach out to us via DM using the 'Send us a Private Message' option below, without including any personal or account information. We'll do our best to assist you in this regard. - Waseem

How does an OTP order get delivered without OTP? My Feb 13 delivery never arrived, yet marked delivered. After 2 days of follow-ups, Amazon now asks me to wait 3 more days. This is not acceptable. @AmazonHelp @amazonIN @amazon

@AmazonHelp @amazonIN @amazon @AmazonHelp I have been following this link...this has already wasted me a lot of time....still the issue is not resolved

@AmazonHelp @amazonIN @amazon @AmazonHelp There is no resolution yet. What a pathetic customer service. Shame @amazonIN @amazon @AmazonHelp

Everyone talks about RAG, but very few explain how real-world search systems actually work. Here’s the 3-layer search architecture behind modern AI products (yes — e-commerce search, support bots, AND RAG apps): - Dense / Semantic Search Understands meaning → finds conceptually similar content even without keyword match. - Sparse / Keyword Search (BM25) Exact term matching → critical for names, errors, IDs, logs, docs. - Re-ranking (ColBERT / Late Interaction) Re-sorts top candidates using token-level relevance → makes results feel “smart.” This Hybrid Search setup is what powers production-grade AI systems today. If you're building RAG or any LLM-based product, this architecture is a game changer. I broke down the full pipeline + demo + code example in a video. youtube.com/watch?v=QuDGpV…

Most people know “semantic search”. Very few understand how to combine it with keyword search + re-ranking. That combo = Hybrid Search and it’s how real systems (e-commerce, support, documentation) deliver accurate results. I just published a simple, step-by-step video breaking it down + showing a real demo. youtube.com/watch?v=QuDGpV…

@Agam_io Yup..time flies

How is it december already ?

@vybhavbhadri @X Let's connect 🙌

Hey @X , I’m looking to connect with people interested in: ✅ DSA, LLD, HLD 🎨 Frontend (React, NextJs) 💼 Backend (Python, NodeJs) 🧠 Gen AI / Data Science 🤖Agentic AI 🎯 Startups Let’s learn & build together 💪

Exploring @qdrant_engine hybrid search today dense vectors + sparse BM25. Trying to really understand how both signals merge to boost retrieval quality. Fun stuff. #buildinpublic

LLM stands for?

RAG is like an AI that checks your notes before answering your question. Search first → answer later → better accuracy. . . . . . . Follow to level up your GenAI skills 🚀