Kontxt
Captures highlights, notes, and reading signals from web pages and documents, turning them into a permission-aware memory that can be retrieved by AI assistants.
Launch screenshots from Product Hunt, May 2022
Kontxt is an enterprise context engine that captures highlights, notes, and reading signals from web pages and documents, turning them into a permission-aware memory. It solves the problem of AI assistants lacking context by providing a governed, retrievable memory layer. The product targets executives, IT, operations, and other B2B teams in mid-market and enterprise companies. It is positioned as a model-agnostic, agent-agnostic layer that sits above the AI stack, delivering permission-checked context to any AI via MCP and API.
Key features
- Capture highlights, notes, and tags
- Passive reading trail tracking
- Permission-aware memory governance
- Ranked hybrid retrieval
- MCP and API integration
- Citation attached to retrieved blocks
- Model-agnostic and agent-agnostic
- Auditable 'why' for each block
Social posts
- No social media activity within the last 30 days
GTM channels
- No GTM activity detected
ICP
- Enterprises
- Mid market companies
- IT teams