Skip to main content

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.

​
Get Started

Set up Kontext by creating your Product, defining the dimensions that matter, configuring AI analysis, and connecting your Feedback sources.

​
1. Create your account

Kontext is currently in early access. Request access to get started. Once approved, you’ll receive a link by email that lets you register.

​
2. Create a Product

A Product in Kontext represents the thing your users give feedback about. It’s the top-level container for all your Feedback, Features, and Problems. The app guides you through this as your first step.

​
3. Set up your reference data

This is the most important part of getting started. Before any Feedback flows in, you need to define the dimensions that Kontext’s analysis pipeline will use to qualify what it detects. These are your Features, Actors, and Contexts — together, they are the reference data that turns raw Feedback into structured, actionable understanding.
Invest time here before connecting any Feedback source. The quality of Kontext’s analysis depends directly on how well these dimensions describe your Product and the people who use it. An LLM can help you draft initial definitions that you then refine.

​
Features

Features are a tree that describes everything your Product is made of — from broad areas down to specific capabilities. The pipeline uses your Feature tree to anchor detected Problems to specific parts of your Product. Without Features, you won’t be able to narrow down where Problems are happening. See Features & Releases for more detail.

​
Actors

Actors are whoever or whatever encounters a Problem — a product manager, an end user, an API consumer, an AI agent. Each Actor has their own perspective on what matters and what’s frustrating. The pipeline tries to identify Actors from what it reads — but without definitions to match against, there’s nothing to attach detections to and no dimension to explore.

​
Contexts

Contexts describe the environment or situation your Product is used in — an enterprise team, an early-stage startup, someone on the go, a regulated industry. The same Problem means very different things depending on the Context. The pipeline tries to identify Contexts from the content and metadata — but without definitions, those detections have nowhere to land. See Contexts & Actors for more detail on both dimensions.

​
4. Configure AI analysis

Without an AI provider configured, Kontext cannot analyze incoming Feedback. You’ll need to set this up before connecting sources.
  1. Go to Product Settings → AI Configuration
  2. Add your API key for your preferred provider (Anthropic, OpenAI, or Ollama)

​
5. Connect your Feedback sources

Kontext integrates with tools where your Feedback already lives. Once connected, Integrations automatically sync new items as they come in — every new issue, ticket, or discussion becomes Feedback for the pipeline. With your Features, Actors, Contexts, and AI provider in place, Kontext will start analyzing everything automatically.
  • GitHub — issues and discussions become Feedback automatically
  • Linear — track issues that map to user Problems
  • Slack — push conversation threads as Feedback
  • Fathom — completed call transcripts become Feedback automatically
Go to Organization Settings → Integrations to connect your first source. Need an Integration we don’t support yet? Let us know — we build new Integrations on demand.