Two years ago I built AGENTS24X7 — an AI SEO-content agent with a WordPress plugin and a Shopify app. The engineering worked. We got exactly one paying user. Investors passed, my co-founder quit, and my brother (yes, my brother) cancelled too.
That failure taught me the only lesson that matters: distribution is the product. Nobody knew we existed.
So this time, I inverted everything.
The Problem We Kept Running Into
I run Codpal, a CTO-as-a-service company. For years, we walked into engineering departments needing to answer one question fast: how is delivery actually going?
The data lived scattered across four different tools — Jira, GitHub, GitLab, and Azure DevOps — and none of them spoke the same language. Commercial platforms exist, but they lock your engineering data behind five-figure enterprise paywalls.
So we built an internal cockpit. It consolidates git activity and issue tracking into one warehouse to surface DORA metrics, team focus, burnout risk, and CapEx vs. OpEx — the whole department’s operational story on a single dashboard. It let us sit with leadership on day one with hard data instead of guesses.
Then we open-sourced it. It’s called Deckgauge — self-hosted, single Docker deploy, FSL-1.1 (converts to Apache-2.0 in 2028).
Live demo: demo.deckgauge.com
GitHub: https://github.com/Codpal-Limited/deckgauge
The Part Builders Will Find Interesting
Deckgauge was largely built by a multi-agent workflow —The tool that measures engineering delivery is itself an experiment in AI-assisted software delivery. (Happy to do a deep-dive post on this setup if anyone wants it.)
What I’m Doing Differently This Time
Building in public: Show HN and Product Hunt launches queued up.
Founder-to-founder outreach: Personalized outreach to engineering leaders and creators rather than generic cold ads.
Open-source core: The self-hosted engine is free and transparent; monetization will live strictly in managed cloud and enterprise governance.
Zero vanity metrics: Documenting the real numbers, especially the ugly ones.
Where We Stand Today
⭐ GitHub stars: ~1 (my mother hasn’t starred it yet, so that’s an honest 1)
💰 Revenue: $0
👥 Design partners: 0
📦 Launch spend so far: $54 ($49 DevHunt + $5 AlternativeTo priority review)
Next post: the raw launch numbers, traffic sources, and conversion fallout — whatever they end up being.
Ask me anything. Especially the critical kind — last time nobody was around to give it.
The previous failure gives you a useful baseline. With Deckgauge, what signal will tell you that outreach is producing real engineering teams rather than just attention?
The main change is in the budget. I make sure not to have monthly expenses, use free tools and free hosting as much as possible
That helps explain the cost side. I’m more curious about the customer signal — what will tell you an outreach lead is actually a real engineering team worth pursuing?