mychance.ai

In Thailand, through a mutual friend, we met a Princeton student who was obsessing over a simple question students and parents ask every year: “Where should I apply—and how can I raise my chances?” We scoped the leanest path to evidence: a two-week proof of concept focused on acceptance-chance estimation using a combination of university data and sentiment signals gathered from public sources.

Our goal was not to “finish a product.” It was to prove the core value loop: input student profile → compute relative chances → show that the signal is meaningful enough to spark traction with stakeholders (students, schools, and investors).

What we built (and only what mattered):

• Acceptance-Chance Engine: Combined university-level quantitative factors with qualitative sentiment to estimate likelihoods.
• Minimal Student Intake: Basic profile inputs to run calculations (kept deliberately short to maximize completion).
• Simple Output: Clear, interpretable results that made the “chance” concept tangible.

No extra bells and whistles. Just the kernel that could be demoed, challenged, and believed.

• Accelerator acceptance: The PoC was compelling enough to get MyChance into an accelerator.
• Investor conversations: The clarity of the problem and tangible results opened doors to early investor interest.
• De-risked direction: The PoC validated that students value both fit and a plan to improve odds, not just a static “yes/no.”

As scope broadened and cost realities shifted, Mirajur chose to continue with a classmate from Princeton for the full build. Our work had done its job: turn the idea into a fundable, credible direction.