The thesis
Careers in Latin America turn on access to people who've already walked the path. MentorHood (mentorhoodlatam.com) is a marketplace that makes that access a product — connecting job seekers with independent mentors, coaches, and experts across professional growth, relocation, and lifestyle, plus a jobs funnel, an education product, and high-touch 1:1 programs. A companion program, MentorOS, helps existing experts package and monetize their experience as structured mentorship.
What I do here
I'm COO and co-founder — the execution side. I run operations day to day: the MentorOS program and curriculum, the mentor cohorts (onboarding, content, session prep), the mentor-verification system, user migration and email nurture, event logistics, and the manual outreach I use to validate a strategy by hand before I automate it. And I build a lot of the internal tooling myself in Claude Code — my marketing approach is openly modeled on the "one-person growth marketer using AI to build their own tools" idea.
What's built
- ◦The platform — a Next.js / React app (Tailwind, TypeScript) on Supabase Postgres, with Clerk for auth, Stripe and Stripe Connect for payments and mentor payouts, and S3 for media.
- ◦"Mentorhood Soul" — the matching engine, live in production. It's a multi-directional system that matches across three datasets — jobs, users, and mentors/products — built as a fleet of n8n workflows: a central router fetches one consistent data snapshot, then fans out to parallel workers (jobs×users, users×mentors, users×products), backed by Supabase RPC functions and AI enrichment.
- ◦AI-enriched jobs — a daily scraping pipeline across several job boards, with AI qualifiers that normalize skills, score remote-friendliness and seniority, infer salary, and suggest the right mentor or product for each posting.
- ◦"Mi Camino" — a unified in-app search experience that surfaces the matches, with personalized email + WhatsApp delivery on a daily cron and instantly at checkout.
- ◦A user-intelligence layer — lead scoring, funnel-stage tagging, and behavioral archetypes derived from a large user export.
Where it stands
A live product with paying users. The matching engine is in production and woven into the platform; the active work is conversion, automation, and growing the cohorts.
Nearby stars
A core Alpicat operating company, built on the same Next.js + Supabase + Clerk + Stripe stack that runs through the whole portfolio. Co-founded with a partner who's also part of GuzPicad.