Founder
Paul Takisaki
AI research methodology applied to scholarship discovery
Paul Takisaki didn’t come from financial aid. He came from AI research. Specifically, he built Ask Three AI, a tool that synthesizes answers from Claude, ChatGPT, and Gemini simultaneously to reduce the hallucinations a single model produces on its own. When his own family started the college search, he applied the same multi-model methodology to scholarship discovery, spent weeks cross-referencing institutional merit policies at target schools, and realized every family needed this. Paul’s background is 20 years at Verizon, starting at a mall kiosk, twelve promotions, and running teams that generated over $1B in revenue. He studied at Bellevue University and the Harvard Program on Negotiations. He is not a certified financial planner. He is not a financial aid officer. He is the guy who applies disciplined AI research methodology to a category that’s still running on Google searches and PDF spreadsheets. That’s the whole credential.
How Ask Three AI led to MeritPlaybook
Ask Three AI exists because any single AI model hallucinates. Claude makes one kind of mistake. ChatGPT makes another. Gemini makes a third. Run the same query through all three at the same time and the disagreements show you exactly where to verify and where to trust. That’s the methodology: use the disagreements as signal. Surface the contradictions, resolve them with a human, and ship the answer that survives the cross-check.
Paul built that tool because he was tired of single-model AI being treated like a search engine instead of a research instrument. The same multi-model verification approach underlies every MeritPlaybook deliverable. When the playbook says a scholarship stacks at a specific school, that claim has been cross-referenced across three models, verified against the school’s own financial aid page, and reviewed by a human analyst before it ships. The methodology came first. MeritPlaybook is what happens when you point that methodology at a category that desperately needs it.
The scholarship research problem
The category runs on the assumption that the bottleneck for families is finding scholarships. It isn’t. The bottleneck is figuring out which scholarships are actually worth pursuing, in what order, against which target schools, and whether each one will displace the institutional aid the student already qualifies for. A 1.5 million-scholarship database doesn’t help with that. The 200 generic awards a free matcher returns don’t help with that. Twenty browser tabs and a Google Doc don’t either.
Paul saw this firsthand when his own family started the college search. Every aid office publishes its policies in PDFs that contradict each other across schools. Every database lists the same low-yield awards in roughly the same order. Every search result is either a thinly-rewritten listicle or a database wrapper. Nobody reconciles the contradictions for one specific student facing one specific list of target schools. So he reconciled them himself. Spent weeks cross-referencing institutional merit policies, building a stacking analysis no scholarship database produces, and shipping it to his own family before he ever thought of charging for it. That document is what MeritPlaybook turned into a product.
Background
Paul spent 20 years at Verizon. He started at a mall kiosk, was promoted twelve times, and led teams of more than 1,000 people running an organization that generated over $1B in revenue across his tenure. He left through a Voluntary Separation Program to ship AI products full-time.
He studied at Bellevue University. He completed the Harvard Program on Negotiations. Today he runs Takisaki Strategy alongside MeritPlaybook and writes about AI research methodology at paultakisaki.com.
What Paul does not claim
The honest version of the credibility story matters more than the marketing version. Paul is an AI research specialist who applied a disciplined multi-model verification methodology to a category that’s still running on Google searches and PDF spreadsheets. That’s a real credential and a real differentiator.
He is not a certified financial planner. He is not a financial aid officer. He doesn’t predict scholarship awards, doesn’t guarantee outcomes, and doesn’t claim to know more about your specific school’s institutional aid policy than the school’s financial aid office does. Every MeritPlaybook playbook is reviewed by a human analyst before it ships. Every claim is sourced. Every scholarship recommendation links back to the school’s own financial aid page so you can verify it yourself. The product is the playbook, not Paul’s resume.
MeritPlaybook delivers the same multi-model research methodology for your student in 72 hours. See a real sample playbook or start your own.