> ## 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. # Core Concepts > The key ideas behind how Kontext works. # Core Concepts Kontext is built around a few key ideas that shape how it processes Feedback and surfaces understanding. ## The problem with summaries Most feedback tools summarize. They compress hundreds of data points into a few bullet points. This is useful for getting a quick overview, but it **flattens the signal** — stripping the distinctions that tell you how a shift in Context changes what you should build, for whom, and what success looks like. Kontext preserves these distinctions by structuring Feedback around **Contexts**, **Actors**, and **Problems** — and maintaining the connections between them. ## Key concepts at a glance | Concept | What it is | Why it matters | | ------------------------------------------------------ | ----------------------------------------------------------- | ------------------------------------------------------------ | | [Features & Releases](/concepts/features-and-releases) | Your Product's Feature tree and how it evolves | Anchors detected Problems to specific areas of your Product | | [Contexts & Actors](/concepts/contexts-and-actors) | The environments and personas behind Feedback | The same Problem means different things for different people | | [Feedback](/concepts/feedback) | Raw input from users — issues, transcripts, tickets | The source material Kontext analyzes | | [Problems](/concepts/problems) | Patterns detected across Feedback | What your users are actually struggling with | | [Contextualizations](/concepts/contextualizations) | A Problem viewed through a specific Context, Actor, or both | The texture that shapes your decisions | These concepts build on each other. Feedback flows in, Problems are detected, and Contextualizations emerge as the system identifies who said what, in what Context.