AI agents won’t fix hiring. What they run on will.
After eight years of research, we rebuilt the hiring stack from first principles: a new candidate-owned interview powering deeper intelligence, native ATS and sourcing, and one agent connecting candidates with opportunities.
The candidate intelligence layer for the AI era.
Stori replaces the 400-word AI-written resume with a 4,000-word candidate-owned interview: deeper evidence companies can search and compare, and candidates can reuse across opportunities.
One system. Not another layer on the stack.
Candidate intelligence, sourcing, ATS, outreach and workflow live together, so Teller understands the whole process instead of jumping between disconnected tools.
Meet Teller. Your agent on either side of the hire.
Hiring or looking, Teller learns from every decision and gets more useful with every interaction: finding people, finding opportunities, and taking on more of the work as trust builds.
Our founding principles.
Eight years of research, distilled into eight principles. Everything Stori does follows from them.
For over sixty years, hiring has relied on the same behavioural competency interview. We went back to first principles and rebuilt it.
By replacing competency-led questioning with structured narrative, Stori reconstructs experience the way people actually remember it.
In most real hires, several people could do the job well. The goal is not to find one mythical “best” person, but to identify who is worth serious consideration.
We never have all the evidence, or enough time to weigh it with certainty. And no assessment can predict the manager, reorganization, market shift, or other factors that will shape future performance.
Hiring is a probabilistic decision, not a predictive one. The goal is not certainty, but better evidence.
Organizations expand competency frameworks as though performance becomes more measurable the more categories they add. But more categories mostly add overlap, not insight. Additional competency labels are often mistaken for additional precision.
Fewer, clearer constructs are more honest than many vague ones.
The first question in any hire is whether someone is capable of work at this level of complexity. General cognitive ability - g - is among the most consistently replicated predictors of performance in behavioural science, and it matters more as the work grows more demanding.
This is not about hiring the smartest person in the room. It is about being capable enough for the work in front of you.
Performance resolves, again and again, to four enduring capabilities: how someone thinks, how disciplined they are, how they communicate, and how they execute. These are not arbitrary categories. Discipline is conscientiousness by another name; the Big Five sits beneath them. Every role needs all four - only the proportions change.
A small number of capabilities does almost all the work.
People remember their careers as stories - the people, decisions and turning points - not as tidy examples of competencies. Behavioral competency interviews ask for named competencies on demand as if memory were filed like an assessment framework. This is the wrong way to get the most out of candidate interviews.
Assessment should align with how memory actually works.
Resumes, interviews, stories, simulations and references - none of these is the thing we are trying to understand. Each is only a window onto the capabilities beneath. A simulation is not valuable because it is a simulation; a story is not valuable because it is a story. They matter only to the extent they reveal how someone thinks, executes, communicates and applies themselves.
Improve the evidence and you improve the judgement - you do not replace it.
Resumes capture history. Interviews capture moments. Assessments capture abstractions.
Candidate Intelligence unifies narrative evidence, structured assessment, and behavioral patterns into one layer that can be searched, compared, and corroborated over time.