Emmanuel Njoroge

Been building almost entirely with AI-assisted tools, and the thing I've found matters most isn't how fast the AI generates code, it's having a real, repeatable way to verify what it actually did before trusting it's correct.

A recent example: I had a security researcher report what looked like a real vulnerability. Turned out the AI-assisted fix from weeks earlier had shipped correctly, but the researcher's finding conflated two different things, the anon key being visible client-side (expected, normal) versus whether it actually granted unauthorized data access (it didn't, verified by literally running real requests against the live endpoint). Took actual empirical testing to tell the difference, not just reading the code and assuming it was fine.

Curious what others have landed on for verifying AI-generated work before it ships. Do you have an actual process, or is it more ad hoc? What's caught you off guard that a quick glance wouldn't have?

A few weeks post-launch here, and I've noticed a real gap between what I assumed would drive growth and what's actually happening. Cold outreach with genuine, specific research has worked better than expected. Directory listings have been mostly quiet. Community threads like this one, oddly, have produced some of the most genuine engagement.

Curious what this looked like for others further along than me. Not looking for generic "do content marketing" advice, more interested in the specific channel that surprised you, either working way better or worse than expected, or a moment you realized you'd been targeting the wrong kind of user entirely.

What actually moved the needle for you, versus what you thought would?

A routine badge-verification failure turned into discovering that ChatGPT, Perplexity, and Claude s web search couldn t actually read our homepage, only 228 characters of it.

Read more: https://medium.com/@emmanuel_111...