Every company on earth tells you what it is doing through its job postings: what it is building, where it is growing, what it pays, what skills it wants. Canaria turns that signal into data. We collect 10B+ observations of job postings, resolve them to 1B+ unique jobs since 2022, and enrich every one with occupation and industry codes, 100K normalized titles, skills and salary. We link them to 29M company profiles, 200M+ business locations in 243 countries and 7,900+ public security series, and sell the result to hedge funds, HR technology companies, researchers and staffing firms.
We are a small team. Nobody here has a fixed lane, and the tools change every month. What does not change is the job: look at what we have, ask why it is not better, build the better version, check that it is actually true, and ship it.
The role
This is not a job with a fixed lane. One week you might be chasing why a customer's posting count dropped 30% last month. The next you are prototyping a new data product from three sources we already have, running an overnight scrape against a site we have never touched, or writing the page that sells it.
We do not care what you studied or whether you have a degree. We care whether you can look at what we have and ask "why don't we do X", build a rough version of X by tomorrow, notice it does not work, try Y, and come back with specific next steps. History, art, biology, economics, trades, military, a failed startup: all fine. We hire for how you think.
What you will do
- Build and ship with whatever tools get it done fastest. You set the goal, split the work, review the result and decide when it is done. You will not be judged on how much you typed.
- Verify what your tools tell you. They will say "fixed" and "all tests pass". You will not read the code. You will check the number, run the query, make the test fail on purpose, and find the cases where it is wrong.
- Diagnose problems in a pipeline that processes a billion records: form a hypothesis, test it, rule things out, ask why until you hit bedrock.
- Prototype new data products. Combine sources we already collect, propose sources we should, and build a first version in days, not quarters.
- Think about the business, not just the ticket: which companies would pay for this, how it should be delivered (files, API, MCP server), what we should scrape next and why.
- Write down what we learn so we do not learn it twice.
- Whatever else the day demands: talking to a customer, sizing a market, checking a sample file before it ships.
Who you are
- An outside-the-box thinker. You do not accept that the way we do something is the way it has to be done. You bring the idea nobody asked for, and you are often right.
- A self-starter. You see a problem and start working it. You would rather attempt, fail and adjust than wait for instructions or a spec.
- A critical thinker. When something does not work you ask why, and you keep asking. You distrust a result until you have seen the evidence.
- Curious about the world. Job postings describe every industry on earth. You find that interesting, and you notice patterns other people walk past.
- A fast learner, without ego. Most of what we do has not been done before, so nobody is an expert. You pick up new tools, data and concepts quickly.
- Ambitious for the company. You think like a founder: you want to know why we do not sell to a new kind of customer, not just why the build failed.
- Dedicated. You finish things. You iterate until it works and then you ask what is next.
- Comfortable with vagueness. Priorities shift when a source breaks or a customer calls. You adapt without losing momentum or attention to detail.
Nice to have (not required)
- You have already built something real on your own: a tool, a dataset, a site, a bot, a business. Show us.
- Comfort reading data: SQL, spreadsheets, anything that turns a table into an answer.
- Exposure to scraping, data pipelines, labor-market data, alternative data for investors, or HR technology.
- Prior startup, research, trade, military or founder background: anywhere you learned to fix what is in front of you with what is on hand.
What we offer
- A ground-floor role at a small, revenue-generating data company with a very large dataset and a very short distance between you and the customer.
- Direct, daily impact: what you ship goes to paying customers, not a backlog.
- Learn across disciplines: data engineering, machine learning, scraping, sales and product, side by side with the founders.
- [Compensation, equity, location and remote policy]