I built an AI mentorship platform because every "find a mentor" advice on the internet is just "go network lol." Here's what happened
Every career advice article ever written:
"Find a mentor."
Great. How?
"Go to networking events. Reach out on LinkedIn. Join communities."
Cool. And if you're the kind of person who'd rather debug a segfault than make small talk at a mixer?
"..."
Yeah. That's what I thought.
The entire mentorship industry has one solution: be extroverted.
That's not a solution. That's a personality requirement disguised as advice.
So I did what any reasonable developer would do when the existing solutions are broken — I ignored them for years and then rage-built my own platform at 2am.
Meet MentorKul.
You ask a question. You get an instant AI-powered answer. Then you get matched with a real human mentor who's actually been where you are — not some random "Top Voice" who posts motivational quotes between golf rounds.
But here's the part I'm unreasonably excited about:
AI Twins.
Every mentor can train their own AI avatar. It learns their thinking, their advice style, their experience. So when you're panicking about a job offer at 11pm on a Sunday — the AI Twin is there. Not a generic chatbot. A version of someone who's actually navigated that exact situation.
Mentorship that doesn't require someone to be awake. Revolutionary? Maybe. Slightly unhinged? Definitely.
The build:
Solo dev. Node.js, React, TypeScript, PostgreSQL. Claude Code did so much heavy lifting that I genuinely felt guilty — like having a group project partner who actually works.
What should've taken 6 months took about 2. The AI didn't make architectural decisions for me, but it absolutely deleted the boring parts. Boilerplate, CRUD, tests — gone. What was left was the stuff that actually requires a brain: semantic matching with embeddings, the AI Twin training pipeline, mentor-mentee routing.
The interesting technical problem: how do you match someone's messy, half-formed career question with the right mentor? Not keyword matching — that gives you garbage. I built a multi-factor scoring system using vector embeddings. Think of it as "Tinder for career advice, but the algorithm actually works and nobody ghosts you."
Where I am now:
Live at mentorkul.com
Small but real user base
$0 revenue (the classic indie hacker starter pack)
Zero marketing budget
One founder who writes better TypeScript than marketing copy
What AI actually changed:
Speed — 2-4x on execution, 0x on judgment. AI can scaffold an entire feature in minutes. It cannot tell you whether that feature should exist.
The risk threshold — A 6-month failed project is devastating. A 2-month experiment is Tuesday. That math changes everything about what you're willing to try.
What AI didn't change — Knowing what to build, who to build it for, and how to get it in front of them. The product is the easy part. Getting humans to care about it is the actual job.
The uncomfortable truth I'm living right now:
I built a platform that connects people with mentors.
My biggest challenge is connecting with people.
The distribution problem is real. I catch myself opening VS Code when I should be writing posts like this one. Shipping code gives you a green checkmark. Posting on the internet gives you existential dread and maybe three likes from bots.
But here I am. Writing this. Because the product actually works, and the people who need it don't know it exists yet. And that's a worse outcome than a cringey post.
What I've learned:
The "find a mentor" problem isn't going anywhere. Every week I see the same questions on Reddit with hundreds of upvotes: How do I find a mentor? Where do I even start? I feel completely lost in my career.
Nobody has a good answer yet. Not LinkedIn. Not networking events. Not the guy charging $500/hour for "executive coaching" that's basically just nodding and saying "what do YOU think?"
I think AI Twins are the unlock — not because AI replaces human connection, but because it removes the calendar and timezone problem. Your mentor lives in Singapore and you're in London? Their AI Twin doesn't care about timezones.
If you've ever stared at the ceiling wondering if you're doing your career completely wrong — with nobody to ask — that's exactly who I built this for.
Now I just need to find you.
For the builders here who've made the jump from shipping to distributing:
What actually cracked it open for you? Not theory — what specific thing worked?
mentorkul.com — free to use, no card required
I know how tough that transition from building to distributing can be. I happen to know a couple of solo founders who made that jump and might be willing to answer your questions about what worked for them for free.
This hit — especially the part about feeling lost even after getting in.
A lot of people don’t talk about that phase, but it’s real.
The AI + human mentor combination is interesting, especially if it can actually feel personal rather than generic.
Curious — right now, do people connect more with the idea of “AI mentor” or the human mentorship side when they first land?
I expected people to gravitate toward the human mentor side. Instant credibility, real experience, the "proper" version of mentorship.
What actually happens: most users try the AI answer first. Not because they trust it more — because it's there right now. The friction of waiting for a human, even a few hours, is enough to kill the moment. Career anxiety has a short half-life. By the time the mentor responds, the user has already convinced themselves to take the job / not quit / send the email.
The AI handles the 11pm panic. The human mentor handles the 3-week strategic conversations. Different jobs, different moments.
The real insight for me has been that "AI vs human" is the wrong framing. It's "now vs later." And users pick now almost every time.
Still early days though. Ask me again in 6 months and I might have a completely different answer.
What are you working on?
That “now vs later” framing is sharp — explains a lot of behavior beyond just mentorship.
I’ve been spending time around early-stage products and one thing that keeps showing up is how much of the outcome is decided in the first few seconds — not just by the product itself, but how it feels immediately.
Two products can solve the same problem, but the one that feels clearer and more “right” upfront gets the user — especially in those high-anxiety moments you mentioned.
So lately I’ve been more focused on that layer — how products are perceived before people even properly experience them.
Still exploring it, but it’s been interesting to see how much it affects conversion and retention without changing the actual product.
Curious how you’re thinking about that on your side — especially for first-time visitors hitting MentorKul.