Companion-device data

Connected to everything.

Connect to any AI agent via MCP. Turn companion-device signals into products with the API. Pipe them into any tool you already run. AiFi handles the audio and the data, you decide what happens next.

The signal you're missing

See your companion-device data on the second screen.

Today the companion device is a black box. AiFi opens it. Every time a viewer connects their phone, picks a feed, boosts dialogue, switches language, or taps a Scene Commerce offer, that's a first-party signal, sent back to you in real time.

Feed it to your ad stack, your analytics, or Scene Commerce, and turn second-screen behaviour into revenue you can measure. It's data you don't have today.

Ad targeting & measurementScene Commerce conversionsFeed & feature engagement
// event streamed back to you
{
  "event": "feed.selected",
  "session": "sd_9f2c...",
  "device": "companion",
  "feed": "match_b",
  "audio": "dialogue_boost",
  "offer_tap": "scene_1042"
}
You own the data, 100%. AiFi returns your companion-device signals to you and keeps no rights to them. It's never resold or shared, you decide where it goes.
Build on AiFi
SDK & API

Build it into your own app.

One SDK for iOS and Android, plus a clean API to start a session, switch feeds, trigger audio, and read the companion-device signals back into your own systems. Ships as software on the audio path, so nothing changes in your video pipeline.

// join a session
POST /v1/session/join
Authorization: Bearer
Talk to our engineers ›
MCP

Ready for your AI agents.

The companion-device signals AiFi returns work with the Model Context Protocol, so the AI assistant your team already runs can read them, with your permissions and data boundaries intact.

  • Secure access for agents and assistants
  • Permissions and data boundaries carry over
  • Traceability back to the original session
Talk to our engineers ›
What you can do with MCP

The Model Context Protocol (MCP) is an open standard that lets an AI assistant read data through tools you connect to it. Point the assistant your team already uses at your AiFi companion-device data, and prompts like these work in plain language, against your real second-screen numbers:

Which feeds did fans switch to most during Sunday's derby?
Show the Scene Commerce offers that converted during the finale.
How many viewers used dialogue boost or audio description this month?
Where did companion-device engagement spike around the ad breaks?

The assistant pulls the sessions it needs, reads their signals and Scene Commerce events, and answers, for a single broadcast when you point it at one event, or across a whole season when you let it find the moments for you. The data is yours; you decide which tools and agents see it.

Where AiFi plugs in
iOS & Android SDK
Drop AiFi into the app you already ship.
Roadmap
REST API
Control sessions and audio from your own systems.
Roadmap
MCP for agents
Give AI agents secure access to AiFi.
Roadmap
HLS & DASH
Works on the manifest you already deliver.
Roadmap
CDN & origin
Sits on the audio path, pipeline-agnostic.
Roadmap
TVU MediaMesh
Companion audio for live production.
Roadmap
Your video player
Pairs with the player you already run.
Roadmap
Webhooks & events
Trigger actions on session events.
Roadmap
Missing one?
Tell us the integration you need first.
Ask us

Everything here is on our roadmap or in active design, built to fit the stack you already run. Tell us which one matters most and we'll prioritise it with you.

Built to fit your stack.

Tell us how your platform is put together, and we'll show you exactly where AiFi connects.

Talk to our team