Glance for AI Agents¶

AI agents — Claude, ChatGPT, Gemini — are becoming the primary shopping surface for a new generation of users. When someone asks an AI "what should I wear to a wedding under ₹3,000?", that's a purchase intent moment. Today, most agents have no answer. Glance AI changes that.

The opportunity¶

Every major AI platform is opening up to third-party integrations via the Model Context Protocol (MCP) — the standard that lets AI agents discover and call external tools the same way a browser discovers web pages. Claude and ChatGPT already support it. More are coming.

The first fashion brand to show up in those assistants owns that intent. We're ready.

What we've built¶

A production-grade MCP server that connects any AI assistant to Glance's fashion catalog in a single integration step.

The partner adds one URL and one API key to their assistant's registry. From that moment, their users can:

Discover products through natural-language conversation — "red dress for a wedding, under ₹3,000"
Refine in real time — "now under ₹2,000", "show me only Zara" — without starting over
See a shoppable carousel rendered directly inside the chat surface, with product images, brand, price, and Buy links
Build outfits — "mix and match this top with something casual" — and get a visual outfit widget inline
Try on virtually — one tap inside the conversation, no app switch required

All of this works inside Claude, ChatGPT, Cursor, and any other assistant that speaks MCP — with no custom code per platform.

Why this is different from a search API¶

A search API returns data. An MCP integration returns an experience.

When a user searches on Google Shopping, they leave their current context, scan a results page, and make a decision in isolation. When a user asks an AI assistant the same question and we're integrated, the assistant becomes the stylist: it asks follow-up questions, remembers what the user already looked at, explains why it picked those items, and renders the carousel right there in the conversation.

We control the rendering. We see which products got attention, not just what was searched. That's the attribution signal that makes fashion recommendations meaningful — and measurable.

Status¶

Capability	Status
Natural-language product search	Live
Multi-query outfit search (parallel)	Live
Shoppable product carousel	Live
Product detail card (tap to expand)	Live
Mix & match outfit builder	Live
Virtual try-on (initiation)	Live
Wishlist save / view	Live
Image upload (visual search, style match)	Live
OAuth 2.0 with Google SSO	Live

The integration ask¶

For a partner to go live:

Add our endpoint to their MCP registry — one URL, one API key
Agree on OAuth scopes (takes one meeting)
We handle everything else: search, ranking, rendering, session state, try-on

Their users get a fashion-aware AI. We get a new discovery surface with full-funnel visibility.