I test AI on real businesses, then publish what holds up
I am Tyler Ang. I run AI tools and new models on real client work, then write up what actually worked here, free and cited. This page is the short version of how I got here and how I judge what is worth your time.
Connect on LinkedInEvery week people ask me which AI tool to use, or whether the new model everyone is posting about is actually any good. Most of the answers online are either launch-day hype or rehashed press releases. Almost nobody has actually run the thing on real work before telling you what to think.
That is the gap this library fills. I spend my days using AI tools and new models on real businesses, the kind of work where a wrong call costs real time and money. When something earns its place, I write up what it is good at, where it breaks, and how it compares, with the sources linked so you can check me.
I learned to work this way the hard way. At 23 I dropped into an e-commerce incubator, managed a brand called Durian Bakery, and learned that perception beats product. I built sportifate.com from zero to 6,800+ organic users in six months with no ad spend, and hit 500K+ TikTok views in under three months. Across 255+ businesses since, the lesson held: test on reality first, then trust the result. This site is that habit, made public.
I find the people who actually know AI, pressure-test what they share, and publish what holds up.
How I got from e-commerce to testing AI on real businesses
Covid changed everything
Started dropshipping on Shopee and Lazada. Could not scale it, but learned marketing is the real leverage.
Das Umbrella e-commerce incubator
Managed Durian Bakery. Learned influencer marketing, affiliate strategies, and how e-commerce brands actually grow.
Built sportifate.com
Built a sports blog to 6.8K+ organic users with zero ad spend. Grew @sportifate TikTok to 500K+ views in under 3 months.
Went into consulting
Started consulting with businesses on digital marketing. Discovered most fail at the offer level, not the tactics.
255+ businesses later
Saw the same pattern across hundreds of businesses, then started testing every new AI tool and model on that real work to see what actually holds up.
How I decide what earns a place in the library
I test before I publish.
Every tool and model here gets run on real client work first. If it did not earn its place in practice, it does not get a writeup. What you read is what actually held up, not what a launch post claimed.
I cite, so you can check.
Claims link back to where they came from. The builders whose work I lean on get credited, the sources get linked, and you can verify everything yourself instead of taking my word for it.
I cover what is new while it is still useful.
When a model drops, the useful read is the early one. I cover new releases fast, then keep the honest comparison up to date as the dust settles and the real strengths and limits show.
Nothing here is for sale.
No paywall, no lead funnel, no upsell. The library exists so the signal about AI stops being buried inside hour-long videos and long threads. You read it, you use it, that is the whole deal.