About
I am not a teacher. I have never run a classroom. I have read this material carefully, and where I think it is shaky, I say so on this site instead of smoothing it over.
That is the honest starting point, not a caveat added at the end. If you came here wondering why a stranger's opinion about your classroom is worth reading, that is the right question, and this is the answer: it might not be, and you should weigh it accordingly.
Who I am
My name is Kyle Zhang. I am a founding engineer at Revola AI, where I build AI systems for a living. I spend most of my working hours deciding which AI capabilities actually hold up in production and which ones only look good in a demo. That is the vantage point this site comes from: not a classroom, but a few years spent watching which claims about AI survive contact with real use.
The consulting business, named plainly
I also run an AI consulting practice at consult.kylemzhang.com, helping small businesses, non-profits, and schools use AI tools without wasting money on the wrong ones. This site is not that. There is nothing to buy here, no service on offer, and no form that turns a reader into a lead. If a school administrator reads a guide here and later becomes a client through the other site, that is a real connection I want disclosed up front, not discovered later.
What pays for this
The consulting work is what funds this site. I write and maintain everything here on my own time, and that time is affordable because client work pays the bills, not because this site earns anything of its own. There is no sponsor, no affiliate link, and no advertising on any page here, and there will not be. Saying plainly where the money comes from is what makes the rest of this disclosure trustworthy - a "just because" version of this page would be easier to write and harder to believe.
Why I am doing it anyway
Most of what I read about AI in schools is written by people selling a product into them, and it reads that way. I wanted somewhere to put the parts of the research and the practical guidance that do not fit neatly into a pitch: the studies that only half hold up, the prompts that work better than the ones in a vendor's demo, the questions a department should ask before it adopts anything. None of that needs a business model attached to be worth writing down, so I kept it separate from the one that has one.
What happens to what I learn
When something changes my mind or teaches me a lesson worth passing on, it turns into a new guide or an update to a research page here, not into anything with a sales pitch attached. I am budgeting about four hours a week for this, mostly early mornings before client work starts, and I would rather say that number out loud than make an open-ended promise I cannot keep.
If a stretch goes by with nothing new added, that is not a sign the project failed. It means there was not anything worth writing down that week. Everything already published stays up either way, and if this site ever goes quiet for good, I will say so here rather than leave the front page asking for conversations I have stopped having.