Hi, I'm Álvaro.

JobIt

2026 · launched

A CV advisor for Chile. It shows what your CV is missing by quoting your own lines, rewrites it without making things up, and tailors it to each job post.

Why

Most CVs in Chile describe the role, not what the person achieved. And when you finish university, nobody explains how to tell your experience.

What I did

I defined the product, the business model and the experience, and built it with Claude Code: diagnosis, rewriting, CV from scratch, CV tailored to each post with a cover letter, application tracking and payments with Mercado Pago.

What I learned

How to take a complete product to production on my own: login, payments, emails, 30 test suites and an AI evaluation with real CVs. And that critical rules, like never inventing numbers, have to be enforced in code; writing them in the prompt isn't enough.

usejobit.comInstagram

Personal finance coaching

Since late 2025 · 10+ clients

I teach people to get their personal finances in order: track their spending, save, and understand the basics of investing, with practical tips to get started.

Why

Many people go from one day to the next to managing their own money, with no tools at all. With what they earn they shouldn't be in trouble, but without order they end up there. I had built a system for my household finances that worked, and wanted to share it.

What I did

Paid coaching in three modules that I usually cover in one or two sessions: budgeting and expense tracking, saving and emergency funds, and an introduction to investing. Each person gets a spreadsheet adapted to their priorities. Almost all clients come through referrals.

Results

Almost all of them made their first savings outside a term deposit, and those who adopted the spreadsheet now track all their expenses.

What I learned

The problem is rarely how much someone earns; it's the lack of order. The spreadsheet is the core of the coaching, but it doesn't suit everyone: some get more out of the practical saving and investing tips, so I adapt each session to the person.

CMR Calculator

Aug 2026 · live

Helps decide whether to redeem your CMR points now or wait for the new format, and shows your break-even point.

Why

Watching a livestream, I noticed people didn't know what to do with their points before the change. It was really a math problem.

What I did

I built it with AI in a few days and checked it against the bank's official communication. More than 10,000 people have used it.

What I learned

It was my first project in production with real users. I learned to ship an application end to end: GitHub repository, automatic deploys on Vercel, domain, social sharing image and analytics to measure usage.

calculadora-cmr.vercel.appGitHub

Flujo Cero

Aug 2026 · personal use

Finds which apartments in Chile actually pay off as an investment, with every number traceable to its source.

Why

I wanted to buy an apartment to rent out, and the published yields didn't survive basic arithmetic.

What I did

A system that every week checks about 2,700 apartments against more than 5,200 real rentals across 384 micro-areas, estimates rent with its own model and emails me a ranked report.

Results

Rankings by district hide gaps of up to 17% between neighboring areas, and many published yields rely on estimated rents with no real comparables.

What I learned

How to build a pricing model with a 5.7% median error and a pipeline that runs on its own. Half the work was cleaning data: catching duplicate listings, seasonal rentals and units already sold.

Mercado Público

Aug 2026 · active

Tools to explore Chilean government purchases and see, case by case, where the State pays different prices for the same product.

Why

People often say public procurement is broken, almost always without showing the data.

What I did

I matched about 23,000 items bought by more than 1,500 public agencies using open data, and built a price comparison tool and a savings estimator.

Results

Better coordinated purchasing could save the State between 1.7% and 4.8%. An official commission reached 1.5–2.5% with a different method. A supposed 51% saving turned out to be mixing different products under one category.

What I learned

How to work with large, messy public data: normalizing products, controlling for region and month, and checking every striking number before trusting it.

Pulso Digital

Jul 2026 · paused

Tracks the social media of 49 Chilean legislators every week, building the history TikTok and Instagram don't show.

Why

Social networks only show a snapshot. The value is in collecting those snapshots over time.

What I did

In one week I went from 3 test accounts to 52 legislators, with 8 metrics (growth, engagement, alerts) and a dashboard comparing government and opposition. It updates itself every week.

What I learned

How to automate data collection with scrapers and GitHub Actions. And that building was the easy part: I paused it to validate whether anyone needs it before going further.