My PM wife was drowning in spreadsheet hell, so I spent 5 months building her a lifeline.

I’m Aminul (you might know me as @aialvi or just “that guy who tweets raw #buildinpublic posts at 3 a.m.”).

This is my third project as an engineer-turned-founder. But this one is personal.

I watched my wife (a Product Manager) live in spreadsheet hell. Customers emailed, tweeted, Slacked, filled forms, and left Intercom notes. All asking for the same 5 things in 47 different ways. She'd spend half her week just merging duplicates and trying to figure out what actually mattered.

I saw her wasting time and getting frustrated, so I decided to fix it.

Every existing feedback tool she tried fixed exactly one problem but created two new ones:

Canny: Beautiful roadmaps, but no decent widget.
Savio / Hellonext: Great deduplication, but no real prioritization frameworks.
Notion + Tally + Airtable: Free, but she aged 5 years trying to maintain that Frankenstein setup.

So in July 2025, I said screw it. I’m building the one tool that does everything we actually need. Nothing more, nothing less.

The "non-sexy" requirements list I started with (still pinned above my desk):

One embeddable widget: Feedback button + live chat + changelog + help center.
Automatic deduplication: AI that actually detects duplicates so she doesn't have to.
Two-way sync: Works with the tools engineers live in (Slack, GitHub, Jira, Linear).
Baked-in Prioritization: RICE, Kano, WSJF (no more spreadsheets).
Lazy-proof Changelogs: AI that writes customer-facing updates from commit messages.

That’s it. No user profiles, no gamification, no “social feed.” Just the stuff that removes pain.

Timeline & Money Spent

Jul 2025: Validated the idea by manually doing the job for 6 beta customers.
Aug 2025: Built the MVP while keeping the beta alive.
Sep 2025: First test users onboarded (all from X & word of mouth).
Oct 2025: Quit freelancing completely to go all in.
Nov 2025 (Today): Launching the MVP to the public.

The Stack (in case you’re curious)

Next.js 15 + Tailwind
Laravel
Postmark for emails
Paddle for payments
OpenAI for deduplication + changelog generation
Vercel + Cloudflare

What actually worked for growth

Shipping in public every single week (even when it was embarrassing).
Giving beta users personal attention and perks.
Writing raw “here’s what broke today” threads. People love the mess.

What’s next

Public API and dev docs (coming in 30 days).
Zapier integrations.
Maybe hire my first employee in 2026 (terrifying thought).

If you’re a founder or PM drowning in feedback chaos, come try FeatureShark.

The first 50 Indie Hackers who sign up with code INDIE50 get 50% off forever (no catch).

Link in first comment 👇

And if you’re building something right now, keep shipping. The mess is normal. The only thing that actually matters is that you don’t stop.

P.S. Yes, I still write changelogs at 2 a.m. sometimes… but now AI does 90% of the work and I just drink coffee and click “publish.”