Hey IH π
I want to share something I've been building in silence for the past few weeks β and why I'm finally talking about it.
Who I am
I'm Harshit. 21 years old. B.Sc. Nursing student, Semester 4, studying 260km away from home in a hostel room.
I run two things alongside college:
Total cash invested in LinkedCraft so far: βΉ1,070 (~$13).
The problem I kept seeing
I work with Indian marketing agencies on their outreach systems. And I kept noticing the same thing:
Most of them manage LinkedIn for 3-10 clients. They generate content using ChatGPT or random tools. They send drafts over WhatsApp. Clients reply 2 days later. They rewrite. Client says "actually first one was better." They miss the optimal posting window. They repeat this every week.
On top of that β every post sounds the same. Generic AI voice. No personality. No pipeline.
LinkedIn was supposed to bring clients. Instead it was just consuming time.
What I built
LinkedCraft β LinkedIn Pipeline System for Indian Agencies.
Not a post generator. A pipeline system.
Here's what's inside:
β 8 post formats β Results Reveal, Hook+Story, Contrarian Take, Case Study, MistakeβLesson, Data Insight, Client Win, Mini Case Study
β 6 Indian tone profiles β Direct No-BS, Desi Grind Story, Data Authority, Founder Raw, Bold Contrarian, Educator Breakdown
β Lead Magnet Pipeline Posts β dedicated mode where you describe your resource, pick a CTA keyword, and get a post that drives "Comment KEYWORD" engagement (8X more reach than link posts)
β From Real Story mode β paste your raw experience, messy numbers and all, get a scroll-stopping post
β Client workspaces β each client has their own tone, niche, post history, and approval queue
β Approval queue β posts sit pending until clients approve in one click. No WhatsApp.
β LinkedIn scheduling β direct OAuth, no Buffer, no third-party apps. Post auto-publishes at scheduled time.
β White-label monthly reports β your agency brand on the cover. Justify every retainer in one PDF.
β Affiliate system β 20% commission, built-in dashboard
Stack: Next.js, Supabase, Vercel, Claude API, Razorpay
My friend Aditya is building the backend. 20-21 days from completion.
Why India-first matters
Every LinkedIn tool in the market is built for US creators.
Taplio = $69/mo. US context. Dollar amounts.
Supergrow = $19/mo. No agency features. No Indian context.
AuthoredUp = No AI. Formatting only.
LinkedCraft:
β βΉ599/mo (~$7) with AI included
β βΉ amounts in every generated post
β Indian niche references built-in
β Only tool with agency client management
10x cheaper than Taplio. The only India-first LinkedIn pipeline tool.
Current numbers (honest)
Waitlist: 47 people
LinkedIn followers: 2,688
First LinkedIn post: 300 views, 138 impressions
Revenue: βΉ0
App status: Backend ~40% done
Launch: July 1, 2026
The distribution plan
No paid ads. No budget. Three engines:
Ambassador program β reaching out to 10-15 LinkedIn ghostwriters. Giving them free Agency+ (βΉ2,999/mo value) in exchange for 2-3 honest posts/month. Just identified Ambassador #1 β Sakshi Dubey, 14,774 followers, Lucknow-based LinkedIn ghostwriter.
Personal distribution β LinkedIn posts 4-5x/week, WhatsApp agency groups, LinkedIn DMs (10/day), Reddit value posts
Free month campaign β post about LinkedCraft on LinkedIn β get 1 month Agency+ free. First 50 only.
What I'm genuinely worried about
Why I'm posting this today
Because building in silence doesn't help anyone β including me.
I want:
And if you manage LinkedIn content for yourself or clients β I'd genuinely love to show you what we're building.
Early access: linkedcraft.in
Launching: July 1, 2026
Will post weekly updates from here.
β Harshit
P.S. The most surreal part: I'm using LinkedCraft to write the LinkedIn posts that promote LinkedCraft. The tool is generating content about itself.
Interesting build.
The thing I'd be careful with is that the biggest launch risk may not be the one you're worried about.
Sometimes a product reaches launch with the technology working, the pricing decided, and the distribution planned, yet still struggles because an earlier decision was never made cleanly.
That's not a call I'd make casually a few weeks before launch.
Hey Aryan,
Thanks for this genuinely useful.
What's the decision you think
might be unclear?
Happy to hear it directly.
Appreciate that.
The reason I'm being careful is that the useful part isn't identifying the decision. It's making the actual call.
That's not something I'd unpack casually in a public thread a few weeks before launch.
If you'd like the tighter version, drop your email and I'll put it together properly.