How can I expose my products to AI agents?

Once your catalog is prepared and semantically searchable, the next step is distribution: making products accessible to AI agents in a way that fits how agentic ecosystems work. In practice, there are two complementary strategies:

1. Publish a feed that agentic platforms can ingest, and
2. Offer an API that agents can query in real time.

ventic.ai supports both approaches.

Publish an ACP-compatible public feed (for discovery and agentic commerce)

For Agentic Commerce Protocol (ACP) style discovery, ventic.ai can publish your product feed in ACP-compliant format through a public endpoint. Example access looks like:

• https://api.ventic.ai/v1/projects/{projectId}/commerce/feed.json

The key detail is that no API key is needed for the public feed. This design matches how agentic commerce expects product feeds to work:

• The feed is intentionally public
• It is meant to be polled and cached by external systems
• It supports HTTP caching via ETag and Cache-Control

What gets exposed (and what does not)
The feed only includes products that are eligible for publication:

• Only items marked enable_search: true are shown
• Only products with status ACP_READY are included
• It does not expose sensitive merchant data like API keys or financials

On the platform side, enabling and controlling public feed access is surfaced in Project Settings → API Keys, with a toggle for public feed availability and a copyable URL.

Why publish this feed
This approach is best when you want:

• ChatGPT-style discovery where the platform can recommend products during conversation
• Compatibility with Stripe ACP flows and “instant checkout” style experiences
• “Multi-platform discovery” where any agent that supports ACP ingestion can read your catalog

In short, the public feed is your distribution channel into ecosystems that operate on ingestion and indexing.

Expose products through the API for real-time agent queries

Feeds are not always ideal. Many agentic use cases need real-time answers like:

• “Show me products under $200 that are in stock”
• “What is the best option for a coffee lover gift”
• “Which models have 16GB RAM”

This is where the API approach shines. ventic.ai provides a Public API Gateway that supports commerce management, semantic product search, knowledge search, and agent chat access.

With this model:

• Agents do not need to download a full feed
• They can query products on demand using semantic search
• They can access a knowledge base (policies, manuals, FAQs) via RAG

To authenticate, you generate an API key in Project Settings → API Keys, then pass it on every request via the x-api-key header.

Build your own agent experience on top of your data

A major advantage of ventic.ai is that it is not just a publishing tool. It is also an agent backend. The platform exposes:

• Search API for products and knowledge with similarity scoring and filters
• Agent module access with provider selection (Ventic, ChatGPT, Claude, Gemini, Perplexity) and session management

That means you can build:

• A shopping assistant inside your site or app
• A support assistant grounded in your policies and manuals
• An internal sales assistant that searches the same catalog your customers see

How to choose the right exposure method

• Use ACP public feed when you want broad discovery through platforms that ingest catalogs.
• Use API-based access when you want controlled, real-time, query-based interactions without distributing the full catalog.
• Use both when you want the widest reach plus the best real-time experience.