About Pathcast

Pathcast is made by Agori Labs.

Why it exists

There is a lot of anxiety about AI and work right now, and most of it lands on people who are one to five years into a career, or still in school deciding on one. The information they get is loud, contradictory, and changes monthly.

The answers on offer are poor. One kind is a free calculator: type in a job title, get back a percentage. "Your job is 63% automatable." That number does not tell you which parts of your week are at risk, which parts hold, or what to do about either. The other kind looks like a calculator but is a bootcamp ad. The free result is designed to make you nervous enough to enrol in a program that costs as much as a used car.

Neither one tells a person what to do on Monday. That is the gap.

What we decided, and why

We score tasks, not titles. A job title is too coarse to say anything useful. In our own research, two roles that share the same O*NET occupation code shared none of their fourteen tasks. So we break a job into the roughly fifteen things a person actually does, score each one, and weight it by how much of the week it takes. The result is a map, not a number. How we score a job explains every step.

You choose the future you believe. Nobody knows how far AI agents get in three years, how fast your particular employer adopts, or what happens to entry-level hiring. We don't either. So we don't assert one forecast. You set three dials for the future you find plausible, and the report re-scores against it. If you change your mind later, so does the report. We think this is more honest than a single prediction stated with false confidence, and it means we can update the underlying capability scores as the technology moves without pretending we saw it coming.

We publish a finite list of fifty roles. Each one was researched by hand: the task list, the credentials, the local market, the non-AI risks, the advice that sounds right for that role and isn't. We could let a model generate a plausible-looking report for any title typed in. We won't. A thin report is worse than no report, because it carries the same authority as a good one. If your role isn't on the list, you can join a waitlist for it, and we'll build it properly or not at all.

We take no money from bootcamps or training providers. Ever. This is the commercial promise that matters. The moment a training company pays for placement, the advice stops being trustworthy, because the "right move" quietly becomes the one someone paid for. Most free career tools are funded this way. Ours is not and will not be. When a report recommends a credential, it is because we checked the provider's own page, confirmed it still exists, and believe it helps. It is never because they paid to be there.

How we make money

We charge the person using the report, and only them. The exposure map, the durable-tasks analysis, three ranked moves and the first two weeks of the plan are free. The full thirteen-week plan and the two 2028 futures cost $29 once. Ongoing re-scoring and alerts are $7 a month. That is the whole model. You are the customer, not the product.

What we haven't done yet

Pathcast is new. We have not tracked outcomes for anyone who followed a plan. We do not have accuracy figures. We have fifty roles, a scoring method we've published in full, and a set of rules about what we won't do. That is what exists today. We'd rather say that plainly than dress it up.

If you think we've got something wrong, tell us: [email protected].