> For the complete documentation index, see [llms.txt](https://herd-security.gitbook.io/herd-security-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://herd-security.gitbook.io/herd-security-docs/trainings/getting-started-with-trainings/individualized-trainings.md).

# Individualized Trainings

## What Are Individualized Trainings?

Individualized trainings make a **conversational training** adapt itself per person. When you turn it on for a training, Herd rewrites each objective's *pass criteria* for the individual learner at the moment they start — using what Herd has observed about how that person works — while the objectives themselves stay the same for everyone.

What stays fixed and what adapts:

* **Fixed** — the number of objectives and their titles. Every learner sees the same objectives, so reporting stays comparable across people.
* **Adapts** — the pass criteria and framing within each objective, tailored to the individual.

Adaptation happens **when the learner starts the training**, not when they're assigned. If anything goes wrong generating a personal plan, the learner simply gets the objectives as authored — individualization never blocks or delays delivery.

{% hint style="info" %}
Individualization is available on **conversational** trainings only, and is configured per training (off by default). You need training-edit access to configure it.
{% endhint %}

***

## Turning It On

**Step 1:** Open a conversational training and go to its **Objectives** tab.

**Step 2:** At the top of the objectives editor, expand the **How this adapts per person** panel (it shows **On** or **Off** at a glance).

**Step 3:** Fill in the required brief — **"How should this training adapt to each person?"** (up to 2,000 characters). Describe *how to adapt, not what to teach* — for example, "Push experienced engineers on secure code review specifics; keep examples basic for non-technical staff."

**Step 4:** Click **Adapt this training per person** to enable it. While it's on, the button reads **Adapting per person**. (If you enable adaptation with an empty brief, Herd seeds a sensible default brief you can edit.)

### What the adaptation draws on

Plans adapt from things the learner has actually *done* — never inferred traits: what Herd has observed about how they work with AI, **quiz questions they've recently gotten wrong** (and trainings whose quiz gate they haven't cleared), and **phishing simulations they've recently failed**. A learner who keeps clicking invoice-fraud lures gets pass criteria that press exactly there.

### Letting people request a plan themselves

Once adaptation is on, a second toggle appears: **Let people request this themselves**. When enabled, anyone in your organization can ask the Herd bot in Slack or Teams for this plan — *"help me get better at AI"*, *"can I get training on X?"* — whatever their scores say.

Self-requested plans are deliberately low-stakes: **voluntary, no due date, no reminders or nudges**, and never counted against the person. In admin views these assignments are labeled **Requested by the learner**. Turning adaptation off also turns self-requests off.

{% hint style="warning" %}
Changes apply to people who **start from now on**. Anyone already in progress keeps the objectives they were given.
{% endhint %}

### Previewing the adaptation

Before relying on it, use the **Preview for** selector to pick an assignee and click **Preview**. Herd generates a sample of what that person would see — it's not saved, just a sample. If Herd has no observed signals for that person yet, the preview tells you they'd get the objectives as written.

***

## Seeing What Each Person Got

On the training's **Results** tab, each assignee row has a **View plan** link that opens **"\<Name>'s training plan"** — the exact objectives that person was given, with a provenance note on what drove the adaptation (for example, *"Adapted using their recent AI conversation weaknesses"*). Each objective shows one of three statuses: **Passed**, **In progress**, or **Not reached yet**.

Objective-level results still roll up correctly across assignees, even when different people saw different pass criteria, because rollups are keyed on the authored objective.

***

## Learner Profiles

Herd can also build a light **learner profile** for each person — inferred from their recent training completions (no new tracking), covering how they pace themselves, when they train, how quickly they start, and how they handle knowledge checks.

{% hint style="info" %}
**Learner profiling is enabled per organization by Herd** and is off by default. Contact your Herd representative to turn it on.
{% endhint %}

When it's enabled:

* **Admins** see a **How they learn** card on a user's **Trainings** tab, with four traits: **Pacing**, **Time of day**, **Time to start**, and **Knowledge checks** — each based on that person's recent completions. Traits a person has corrected themselves are marked "said so themselves."
* **Employees** can ask the Herd bot in Slack or Teams what it has observed about how they learn, and correct it — corrections are recorded alongside (never overwriting) the observed value.
* **Delivery adapts** — a person's preferred time of day can shift queued training sends *later* within your business-hours window (never earlier, and never outside the window), and overdue-training reminders pace themselves to how quickly the person usually starts: quick starters are reminded a business day sooner, slow starters a day later. Your configured reminder cadence is otherwise unchanged.