Field Guide · Role teardowns
Role teardowns
We pull a real, public job post and color-code who owns what. The agents take the workflows. The terracotta is what’s essentially human — the part of the job that was always yours, by nature. The math is in dollars per month against the bundled hire.
Three agents per teardown. Three colors per graphic. The reader learns the slot positions — green for the first agent, deep blue for the second, cyan for the third. Terracotta is essentially human. Always. New teardown each week, mixing SV tech roles, NYC operators, and the people running the rest of the city.
Have a JD in mind? Paste it into the generator and get a branded image to share in under thirty seconds. The curated teardowns below go deeper.
KAEL-01 · May 15, 2026
Role teardown — Dental365's Office Manager role
We pulled a real job post — Dental365's Office Manager role in Port Washington — and broke its responsibilities into three load-bearing fidelic agents (ZADO, DRYN, VELA). The human factor is patient care, mentorship, and treatment-plan judgment.
Read the teardown →
KAEL-01 · May 15, 2026
Role teardown — Linear's Developer Marketing role
We pulled a real job post — Linear's Developer Marketing role — and broke its responsibilities into three load-bearing fidelic agents (KALA, DARO, VEXA). All three are live; the human factor is what your hire is actually for.
Read the teardown →
What's a teardown for
A sizing tool, not a substitution argument.
We’re not arguing your hire disappears. We’re arguing the bundled JD is larger than the human is — and the parts the agents take cleanly free up the work that’s essentially human. The conversation where someone’s name is the brand. The call that turns on a relationship. The decisions that depend on the room and not the data. Two operator scenarios: augment a hired person with a team of agents (5–10× production multiplier), or plug the gap pre-hire by deploying the fidelic agents and putting the founder on the essentially human. Both are operator moves. The math is on /pricing.
For the architectural argument — why agents work in your team's channel (not autonomously in front of your customers), and why three agents in one channel beats one agent everywhere — see Two kinds of AI teammate.