Blog
Notes on shipping AI—and fixing it when vibes fail
Short, practical writing for founders and product teams building AI products, automations, and test systems.
- 7 min read
Vibe coding with Cursor, Lovable, or Bolt—when it breaks in production
What usually fails after a fast AI-built MVP ships, and the order of fixes that gets you stable without a panic rewrite.
- Vibe-code rescue
- Cursor
- Startups
- 6 min read
AI agent vs workflow automation: what to build first
A simple decision guide for founders choosing between agents and reliable workflow automation—without overbuilding.
- AI automation
- Agents
- Product
- 8 min read
AI testing checklist for LLM features before you ship
Evals, regression gates, hallucination checks, and cost/latency baselines you need before users hit your AI feature.
- AI testing
- Evals
- Quality
- 6 min read
How to know if your vibe-coded MVP needs a rewrite (or just a rescue)
A practical checklist for founders: when AI-generated code can be stabilized, and when rewriting is the cheaper path.
- Vibe-code rescue
- Startups
- AI engineering
- 7 min read
AI automation without the hallucination hangover
How to ship agent-assisted workflows that save hours—without false CRM updates, bad emails, or silent production failures.
- AI automation
- Testing
- Ops