Viktor.comAn AI coworker that actually does the work
My College Gate
Free AI college counseling for every student, forever.
20 followers
Free AI college counseling for every student, forever.
20 followers
1.7 million students have no access to a college counselor. Private counselors charge $300/hr. MyCollegeGate is the free alternative, an AI college counselor named Maya, available 24/7. Build your college list, write and review essays, compare your stats to real verified applicants, find scholarships, and get a personalized roadmap to your dream school. Built for first-gen and low-income students. No subscriptions, no paywalls, no data selling. Free for every student, forever. mycollegegate.com
the verification-by-decision-letter piece is what separates this from a chatgpt wrapper, most peer comparison tools just take self-reported stats at face value. genuine question on sustainability though: free forever, no ads, no subscriptions, funded by donations - a 24/7 personalized chat counselor for potentially 1.7 million students is a real LLM API bill at scale. what's the actual plan for keeping this free once usage grows past what donations alone can cover, or is that a problem you're deliberately punting on until it's real?
@galdayan Great question and the right one to ask.
Donations keep it alive now. The long-term plan is institutional sponsorship: colleges pay for a sponsored portal within MyCollegeGate, where students exploring that school receive tailored guidance from Maya. The college gets qualified, prepared applicants instead of cold leads from a name list. The student experience stays completely free.
The math works: a small number of sponsored colleges covers the infrastructure for a large number of students. API costs scale, but so does the value to institutions.
I don't have paying sponsors yet; I'm early. But the architecture is intentional: free product builds the user base, the user base makes the sponsorship credible, and sponsorship makes free sustainable.
If I'm wrong, I'll figure out the next model. But I won't add ads, and I won't sell data. Those are lines, not preferences.
@gonzalo_perez1 "those are lines, not preferences" is a good way to put it, and honestly a bootstrap-on-donations-then-earn-sponsorship path makes more sense than trying to raise VC money for a free product from day one. good luck getting the first sponsor - that first logo is always the hardest one to land.
A feature that shows nearby students who matched your profile and got accepted to the same schools would be huge. Real examples from people in your city or zip code make the data feel way more useful than national averages, and it helps first-gen students realize they are not alone in this process.
@taagzahide13498 Love this idea, the "you're not alone" angle is exactly right. A first-gen student in rural Texas seeing that someone from their area with similar stats got into a great school is way more motivating than a national average.
Adding this to the roadmap. The honest limitation right now is we'd need students to optionally share their region or state during onboarding (not exact location, privacy matters) so we can build enough local data to make the matches meaningful. With a thin dataset showing "0 students near you" would feel worse than not having the feature at all.
But the vision is right, this is the kind of thing that turns data into motivation. Appreciate the suggestion, genuinely.
How does Maya actually pull the verified applicant data you mention? Is that from real submissions, third-party sources like the Common Data Set, or something else entirely?
@kaltal24301 Three layers of data right now:
Real student submissions: students submit their outcomes (accepted/denied/waitlisted) and upload their actual decision letter for AI verification. If the document checks out, the result goes live in our community dataset. If it's flagged, it goes to manual review. This is the core of what makes the peer comparison real.
College-level stats, acceptance rates, GPA ranges, SAT/ACT percentiles, and financial aid data populated from publicly available Common Data Set and institutional reporting.
Consenting to community data, we reached out to students who publicly shared their results and got explicit permission to include their anonymized stats. No scraping, no pulling data without consent.
The verified submission layer is what I think makes this different — most comparison tools rely on self-reported stats with no verification. Anyone can claim they got into Harvard with a 3.2. On MyCollegeGate, you have to upload proof.
The honest limitation: we're early, so some schools have thin data. We show sample sizes on everything so students know when a pattern is based on 50 verified results vs 5. Transparency over false confidence.
How does Maya actually compare my stats to real verified applicants without that data getting sold or shared with colleges down the line?
@dogrucamer51928 Short answer: we never see your name attached to your stats, and there's no infrastructure to sell it even if we wanted to.
Here's how it works: when you use the chance calculator or peer comparison, Maya compares your profile against anonymized AdmissionResult records in our database. Those records have GPA, SAT, ECs, and outcome, but no names, no emails, no school identifiers tied to them. Your own profile data is tied to your account but never exposed to other users, never shared externally, and never packaged for colleges.
We don't run ads. We don't have analytics tracking. We don't have a lead-gen pipeline. There's literally no mechanism built to sell data because we intentionally never built one. The app is funded by voluntary donations, that's it.
We also have a plain-language privacy page on the site that says all of this in non-legal terms because I think students deserve to actually understand what happens to their data, not decode 40 pages of fine print.
The cynical answer to "how do I know you won't sell it later" is, you don't, you have to trust me. But I built this specifically because I was frustrated by platforms that exploit students. Building the thing I'm against would defeat the entire point.
Curious how Maya handles essay feedback compared to a real human counselor, is it more surface-level grammar stuff or does it actually push back on the story and structure?
@azra1143897 Great question. Maya doesn't do surface-level grammar corrections. She evaluates on a rubric that covers five areas: personal voice and authenticity, organization and flow, depth of reflection, specificity and detail, and mechanics. Each gets its own score out of 10 with specific feedback that references actual lines from your essay.
The part I'm most proud of is the honest verdict, Maya will straight up tell you "this essay is not ready to submit" if it isn't. She'll tell you your opening is generic, your reflection is shallow, or your essay sounds AI-generated. No "great job!" when it's not.
She also pushes back differently for first-gen students specifically, a lot of first-gen students undersell their background or apologize for their circumstances in essays. Maya catches that and reframes it as the strength it actually is.
That said, we also link to Get Schooled for free human essay review (7-10 days turnaround) because AI and human feedback together is better than either alone.
Would love for you to try it and tell me if the feedback feels real enough — that's the thing I'm constantly trying to improve.
How does Maya actually get to know a student's strengths and goals well enough to give solid advice, especially compared to a human counselor who can read body language and ask follow-ups in real time?
@selim9xif Honestly, she can't fully replace that. A human counselor who knows you personally, reads your body language, and has years of relationship with you will always have an edge in certain ways. I had that at my school, and I'd never claim Maya matches it.
What Maya does differently is start with your complete profile before the first message. When you go through onboarding, you enter your GPA, test scores, APs, extracurriculars, awards, dream school, intended major, and whether you're first-gen or low-income. Maya reads all of that before she says a word to you, so her first message is already personalized, not "tell me about yourself."
She also asks follow-ups in real time. The essay writing guide specifically works as a back-and-forth interview, Maya asks you one question at a time about your life, pushes for specificity when your answers are vague, and builds your essay from what you actually said in your own voice.
Where she genuinely outperforms most school counselors (not the great private school ones, but the ones serving 500+ students): she's available at 11pm on a Sunday when a student is panicking about a deadline, she remembers everything from previous conversations, and she has real verified admission data to back up her advice instead of intuition.
The gap I'm trying to close isn't "replace amazing counselors." It's "give something real to students who currently have nothing."