I run multiple $10K MRR companies on a $20/month tech stack

Article Summary

Steve Hanov describes running multiple six-figure monthly revenue businesses using an intentionally minimal tech stack: a $5-10/month VPS running Go backends, SQLite with write-ahead logging, local GPU-based AI processing, and strategic use of OpenRouter and GitHub Copilot. He argues this lean approach provides “the exact same runway as getting a million dollars in funding” while maintaining architectural simplicity and freedom to find product-market fit without investor pressure.

Discussion

• Many commenters debated SQLite vs. PostgreSQL for concurrent workloads — consensus leaned toward Postgres on localhost via Unix sockets as a pragmatic middle ground that avoids the TCP overhead argument while gaining enterprise features.
• A recurring theme was that complexity like Kubernetes and serverless is often “learned helplessness” driven by cloud vendor marketing rather than genuine technical necessity for most indie projects.
• Several commenters pointed out the real challenge is not the tech stack but finding a valuable problem and reaching customers — the stack is largely irrelevant to actual business success.
• Critics noted the article reads as generic advice wrapped in AI-generated prose, lacking concrete cost comparisons between the recommended stack vs. alternatives like Python + Postgres.
• Practical tips shared: Hetzner auction servers (~€40/month) with Proxmox for running multiple VMs, Fly.io scaling-to-zero with Litestream for $0.10/month hobby projects, and Tailscale + IP whitelisting for SSH hardening.

Discuss on HN