Our beliefs about the future of technical work:
We believe that as agents become more capable, two themes are emerging:
The builder stack collapses
•
Coding agents provide the ability to move across domains in ways that weren’t possible before
•
You can offload more decision making to the agents inside your product meaning the lines between ML work and eng work start to disappear.
•
Approx 20% of product engineering at Basis is teaching agents to tackle non-deterministic workflows. We see that being 70% by end of year.
The distinctions between frontend vs. backend, infra vs. product, ml vs. eng are already melting away.
Engineering > Coding Anything that can be clearly defined and verified will be delegated to an agent. What remains is deciding what to build, breaking down problems from first principles, and designing systems that compound.
The kind of engineer who does well here
We hire individuals, not resumes. Given how fast things are changing, it’s more important to hire the best people than to hire the person with specific experience.
First principles and systems thinking. Many of the problems we solve are net new. There’s no playbook. We want people who can decompose a novel problem, reason from fundamentals rather than pattern match, and design things that compound.
Thinking beyond the engineering. The engineers who do best here understand the domain deeply: the data systems, the workflows, the work our agents are performing. That understanding is what lets you design better systems, train better agents, and make better product decisions.
Ownership of outcomes. Every project at Basis has a single Responsible Party (RP) who is accountable for whether it ships and whether it works. There are always more things to own than people to own them. Engineers own their own work at Basis. Read more about our RP system here
Agency. Having the initiative to pick up the context you need to make good decisions. Talking to the right people, reading the data, understanding the problem deeply enough to know what matters.