> ## Documentation Index
> Fetch the complete documentation index at: https://docs.exterview.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# The Talent Graph
> The shared language of skills, evidence and outcomes that both Exterview and Gradientflo read and write — the reason the two applications are one platform rather than two tools.
**Status: Upcoming.** The Talent Graph is not a product surface today, and
Gradientflo is Upcoming. What works today: Exterview's evaluations capture
skills and evidence against your roles. This page describes the design.
## Purpose
The **Talent Graph** is your organization's private map of skills, evidence and outcomes: what capabilities your roles require, what evidence you have that a person holds them, and what happened after you decided they did.
It is populated by the work you already do. Every role designed, every interview conducted, every evaluation completed and every outcome recorded adds to it.
**The graph is yours, and it is private to your organization.** It is not
pooled, not shared between customers, and not used to train models for anyone
else.
## Why it matters
Most organizations describe the same capability four different ways: one in the job description, another in the interview, a third in the onboarding plan, and a fourth in the performance review. Nothing connects, so nothing compounds.
The Talent Graph makes it one vocabulary across the whole employee lifecycle.
```mermaid theme={null}
%%{init: {'theme':'base','themeVariables':{'primaryColor':'#EEF2FF','primaryBorderColor':'#4B6CFE','primaryTextColor':'#0F172A','lineColor':'#64748B','clusterBkg':'#F8FAFC','clusterBorder':'#E2E8F0'}}}%%
flowchart LR
P([Plan a role]) --> H([Evaluate a candidate])
H -.-> D(["Develop the hire (Upcoming)"])
D -.-> R(["Review performance (Upcoming)"])
R -.evidence.-> G(["Talent Graph (Upcoming)"])
P -.-> G
H -.-> G
D -.-> G
G -.same skills.-> P
```
*Planning a role and evaluating candidates work today in Exterview. Dotted lines are Upcoming: the post-hire steps and the consolidated Talent Graph.*
A capability means the same thing in a job description, an interview, a development plan and a review.
A skill is attached to what actually demonstrated it, not to a self-rating on
a form.
Someone's capability profile updates as they are evaluated, developed and
reviewed, instead of being rebuilt from scratch each time.
Every evaluation makes the next one sharper, because outcomes feed back into what good looks like for your organization specifically.
## How the two applications use it
| Application | Reads it to | Writes to it |
| --------------- | ----------------------------------------------------------------- | ------------------------------------------------------ |
| **Exterview** | Define role requirements and evaluate candidates against them | Evidence from every evaluation, and the hiring outcome |
| **Gradientflo** | Understand what a person was hired for, and against what standard | Ramp signals, development progress and review outcomes |
Exterview owns the graph; Gradientflo reads and writes it. One graph, several applications — the same relationship an organizational graph has to the products built above it.
## What it makes possible
Written against the criteria the person was hired on, not invented by whoever had time.
Where real outcomes are known, they sharpen what your organization treats as
evidence of a strong hire.
Matching an existing employee to an open role using what you already know
about them.
Stating a workforce need as a skill demand rather than a headcount number.
## Governance
The graph holds evidence about people, so it carries the platform's controls without exception.
No access across customers, in any direction.
Reading the graph never exceeds the permissions of the individual asking.
Every entry traces back to what produced it, so any conclusion can be
inspected.
As designed, retention and erasure follow the policy you set. See [Privacy & Residency](/trust/privacy-residency).
## FAQs
No. It is private to your organization and is not pooled or used to train models for anyone else.
No. The graph builds from the work you already do. You can bring your own
framework if you have one.
Skills and evidence are captured through Exterview's evaluations in production today. The consolidated graph surface, and Gradientflo's use of it, are Upcoming.
## Related
*
[Gradientflo](/gradientflo/overview).
*
[Applications on Smaya](/smaya/applications).
[Applications on Smaya](/smaya/applications).