The first rung of the ladder is harder to reach.
Companies are hiring fewer junior engineers, while AI handles more of the simple work juniors used to learn from. Take away enough of the bottom step and the industry has a senior pipeline problem.

We exist to rebuild the first rung.
Blacksmith Experience is an AI engineering team you can join today. Eleven named AI engineers hand you work, read every line you write, and leave behind a record anyone can check. The learning runs on the same rails: a track to what a senior engineer in your field is expected to understand, taught from the code you actually wrote rather than from examples someone made up.
What we are trying to make true is narrow. Someone who can already write code, and has not yet been hired, works through real engineering cycles here with serious review behind them, and leaves with a record that answers the question an employer actually asks: can this person do the work?
This is not a feeling about the market.
Three independent studies, all pointing the same way. Every figure below is linked to its source. Check them.
Early-career employment in AI-exposed occupations
Unemployment among computer science graduates
Share of big-tech hires who are new graduates
Early-career workers are taking the concentrated hit.
Since generative AI adoption, early-career workers aged 22–25 in the most AI-exposed occupations saw a 16% relative employment decline, while more experienced workers stayed stable or grew.
Stanford Digital Economy Lab, 2025 ↗New graduates are vanishing from the hiring mix.
New-grad hiring at the 15 largest tech firms is down more than half since 2019.
SignalFire State of Tech Talent, 2025 ↗A computing degree no longer guarantees an easy first job.
The New York Fed major table reported elevated unemployment for recent computer science and computer engineering graduates.
Federal Reserve Bank of New York, 2025 ↗Seniors were always made out of juniors: bad code, hard reviews, and someone senior staying in the room.
The companies asking for AI-literate senior engineers still need a way to make them. If juniors stop getting reviewed, trusted, and stretched, the senior pipeline thins behind them.

The experience employers already expect you to have.
Not a course with a project attached. An AI-led working week: you bring the code, eleven engineers assign it, review it, and refuse to sign it off until you can explain it.
Four things we hold to while we do it.
Experience cannot be faked, so we do not fake it.
You work on real software. The proof that comes out the other side is proof of real work, not a score from an invented test.
A certificate is only worth the evidence behind it.
Every level points at a specific change, review or decision. Anyone can open it on GitHub and check, or it does not count.
The missing step is everyone’s problem.
If the industry trains fewer juniors, it eventually has fewer seniors. Rebuilding the bottom of the ladder is not charity; it is supply.
Built for the world, starting where it is hardest.
Where entry-level jobs are scarce locally, a real way in matters most. We start there, and we keep it free to begin.
Senior is not a title. It is evidence that survives review.
Free to start, built for the world, so the next generation of senior engineers has a way onto the ladder.
