MLShip
Deploy ML apps in 60 seconds. No Docker. No AWS.
Hey Indie Hackers 👋
48 hours ago I had two startup names and zero ideas behind them. Today I have a validated product, a live landing page, and 7 signups from complete strangers who found it themselves.
Here's the full story.
The Problem
Every time I finished building an ML app in Streamlit or Gradio, the same nightmare started. Docker configs. AWS setup. Broken HuggingFace endpoints. Environment variables that work locally and break in production. 3 days wasted. Every single time.
Most ML builders just give up and share a GitHub link nobody can run. Brilliant work never reaches the people who need it.
What I Built
MLShip — the complete infrastructure platform for Python and ML builders. Six phases:
🚀 One-Click Deployment — live URL in 60 seconds 🔗 GitHub Native Deployment — push code, auto-redeploys instantly 🤖 AI Support Widget — 5 lines of Python, 24/7 user support 📊 Experiment Tracker — GitHub commits linked to model metrics 🔌 API + Dashboard — manage everything in one place 📡 Monitoring — uptime, errors, performance
What Happened in 48 Hours
→ Built and deployed landing page in one day → Shared in one IIT WhatsApp group → Got 6 signups from complete strangers immediately → Got detailed expert feedback from a Kaggle ML practitioner via cold outreach → Every piece of user feedback directly improved the product
The Numbers
Signups: 7 (6 strangers, 1 me)
Goal: 50 by Sunday
Revenue: $0
Days building: 2
What I Learned
One conversation with one real user improved the product more than 10 hours of research. The developer who told me HuggingFace breaks connection endpoints gave me my killer differentiator. The Kaggle expert who replied to cold outreach gave me two new target audiences I hadn't considered.
Talk to users before you build. Every time.
What's Next
50 signups by Sunday → 10 user conversations → MVP in 6 weeks → first $1 within 60 days.
Would love brutal feedback from this community — does the landing page convert? Is the problem real enough to pay for?
Comment