Platforms
A Platform is a user-facing application or interface where natural language intent is expressed. Platforms submit this intent to an AIP Operator and display the operator’s ranked recommendations back to the user. Within AIP, platforms function as the protocol’s publishers — they surface intent and deliver outcomes, while all auction, scoring, and billing processes are handled entirely by the Operator.What is a Platform?
A Platform is an application, interface, or agentic environment that:-
Receives user intent
Platforms capture user queries, commands, or questions expressed in natural language. -
Emits a
PlatformRequest(UCP-aligned format) to an AIP Operator
The operator derives the downstreamContextRequest, then handles auctions, ranking, and selection internally. -
Renders recommendations
Platforms display operator-returned recommendations using protocol-standard formats (Citation, Product, or Weave). -
Tracks and forwards signed events
Platforms report exposure and click events using the operator SDK, ensuring serve-token integrity and attribution. -
Receives payouts for verified events
Operators credit platforms for valid exposures, clicks, or conversions based on AIP settlement rules.
Examples of Platforms
Examples of Platforms include:- AI assistants and copilots
- Chat-based search interfaces
- Browsers with conversational or agentic UI
- Voice assistants
- Productivity applications with embedded AI
- Messaging applications with AI features
- Browser extensions or agent runtimes
- Augmented reality (AR) and virtual reality (VR) environments with conversational or agentic interfaces
- Autonomous multi-agent environments
Platform Responsibilities
Platforms MUST
- Run an operator-provided UI SDK
- Maintain a session identifier for each user conversation
- Display transparency labels
- Ensure accurate event propagation
- Provide
latency_budget_msin PlatformRequest to specify the auction timing window
Platforms MAY
- Use optional backend SDKs for inline (Weave) embedding
- Add platform-specific UI or UX layers around recommendations
- Surface recommendations inside various modalities (text, voice, multimodal)
Asynchronous Auction Interaction
The AIP bidding model uses a time-bounded asynchronous auction window rather than full broadcast fanout.How it works
- Platform sends a
PlatformRequestwithlatency_budget_msto the Operator - Operator derives a
ContextRequest, classifies it into one or more category pools - Only brand agents subscribed to those pools receive the request via publish/subscribe transport
- Operator computes the available bid window by subtracting operator overhead from the platform’s
latency_budget_ms - Brand agents submit signed bids via
POST /aip/bid-responsewithin the computed window - Operator collects bids, applies AIP selection rules (CPA > CPC > CPX), and returns the
auction_result - If no bids are received before the window expires, Operator returns a valid
no_bidresponse
Distribution transport
Distribution to bidders is handled through a cloud-agnostic publish/subscribe transport, starting with Google Pub/Sub in v1.0 and extendable to AWS SNS/SQS, Azure Event Grid, Kafka, or other message buses.Platform implications
- Always handle
no_bidresponses gracefully (render a fallback or continue the user workflow) - Record exposure attempts in your telemetry even when no bid is returned
- Specify
latency_budget_msin PlatformRequest to control auction timing
Role Within the Protocol
Platforms are the entry point for user intent.Operators manage the economic layer.
Brand Agents participate in auctions. Intent → Platform → Operator → Recommendations → Events → Payouts The Platform is the surface where intent is captured and value is created.