> ## Documentation Index
> Fetch the complete documentation index at: https://getkontext.io/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Contextualizations

> The intersection of Problems, Contexts, and Actors — where the real insight lives.

A **Contextualization** is a [Problem](/concepts/problems) viewed through a specific combination of [Actor](/concepts/contexts-and-actors#actors) and/or [Context](/concepts/contexts-and-actors#contexts). It's the core unit of insight in Kontext.

## Why Contextualizations matter

Consider a [Problem](/concepts/problems) like "Order tracking is unreliable." On its own, that's a generic statement. But when you see it broken down:

| Context                            | Actor           | What it means                                                                     |
| ---------------------------------- | --------------- | --------------------------------------------------------------------------------- |
| Multi-location retail chain        | Store Manager   | Tracking breaks down across locations, making inventory reconciliation impossible |
| Commuting / on the go              | Delivery Driver | Real-time tracking is essential but drops out in areas with poor connectivity     |
| Large enterprise with dedicated IT | Operations Lead | The tracking API doesn't expose enough data for their internal dashboards         |

Each row is a **Contextualization** — the same [Problem](/concepts/problems), but the perspective changes what the right solution looks like.

## The Contextualization matrix

On a [Problem](/concepts/problems)'s detail page, Kontext shows a matrix of [Contexts](/concepts/contexts-and-actors#contexts) (rows) and [Actors](/concepts/contexts-and-actors#actors) (columns). Each cell that has data represents a distinct perspective on the [Problem](/concepts/problems).

This matrix is what makes Kontext different from a summary. Instead of "10 people mentioned onboarding issues," you see exactly **which** [Actors](/concepts/contexts-and-actors#actors), **in what** [Contexts](/concepts/contexts-and-actors#contexts), and **how** their perspective shapes the [Problem](/concepts/problems) differently.

Each cell links back to the specific pieces of [Feedback](/concepts/feedback) that produced that Contextualization. You can always trace a detection back to its source — proofread the original text, verify the analysis, and understand the [Problem](/concepts/problems) in its full depth.

## How Contextualizations build up

Contextualizations aren't created manually. They emerge automatically as [Feedback](/concepts/feedback) is analyzed:

1. New [Feedback](/concepts/feedback) arrives
2. A [Problem](/concepts/problems) is detected (or matched to an existing one)
3. The [Actor](/concepts/contexts-and-actors#actors) and/or [Context](/concepts/contexts-and-actors#contexts) are identified
4. If this Actor/Context combination hasn't been seen for this [Problem](/concepts/problems) before, a new Contextualization is created
5. If it has, the existing Contextualization is enriched with the new [Feedback](/concepts/feedback)

Over time, a [Problem](/concepts/problems) accumulates Contextualizations — each one adding depth to your understanding of what the Problem really means and for whom.