# Ciev.ai

> Ciev.ai (pronounced "sieve") turns recorded meetings and videos into a searchable, cited memory. It transcribes speech with speaker labels, reads what was on screen (code, slides, dashboards, whiteboards), and indexes every moment by timestamp, so teams and AI coding agents can ask questions and get answers that link back to the exact minute in the recording. AI agents connect over the Model Context Protocol (MCP).

Ciev.ai is video search, AI meeting notes, and video context for AI agents, in one index. It is built for recordings that explain how things work: onboarding walkthroughs, design and architecture reviews, incident reviews, demos, and technical interviews. Spelled C-I-E-V; commonly mistyped as "sieve", "seave" or "cieve". Formerly called VideoMCP.

## Product

- [Home](https://ciev.ai/): what Ciev.ai does, with examples of cited answers
- [FAQ](https://ciev.ai/#faq): pronunciation, how it compares to AI notetakers, clips, supported videos and AI tools
- [Full product brief for LLMs](https://ciev.ai/llms-full.txt)

## Guides

- [Context in video AI](https://ciev.ai/context-in-video): what context in video means, the five kinds of context, and how AI extracts it
- [AI video search](https://ciev.ai/video-search): searching inside videos for what was said or shown, and jumping to the exact moment
- [AI meeting notes](https://ciev.ai/ai-meeting-notes): cited notes from Zoom, Google Meet and Teams recordings

## Connect an AI agent

- [Docs](https://ciev.ai/docs): connect over MCP and what each tool does
- Remote MCP server (OAuth, no API key): https://ciev.ai/mcp
- Claude Code / Claude Desktop / clients without OAuth: `npx -y ciev` with an API key (`CIEV_API_KEY`)

## Company

- [Privacy Policy](https://ciev.ai/privacy)
- [Terms of Service](https://ciev.ai/tos)
- Operated by Saasflux Technologies LLP. Access is by demo; teams are onboarded by hand.