Getting Started

Go to vritantai.com/signup, enter your email, password, and store URL. Choose your platform (Shopify or WooCommerce).

No credit card required for the 14-day free trial.

Shopify. Go to your Shopify Admin, then Settings, then Notifications, then Webhooks. Add a webhook pointing to your VritantAI endpoint:

https://api.vritantai.com/api/v1/integrations/shopify/webhook
HMAC secret: (shown in your VritantAI dashboard → Settings → Integrations)

WooCommerce. Install the VritantAI plugin from the .zip in your dashboard, then connect it with your API key:

1. Download the VritantAI plugin .zip from your VritantAI dashboard (Settings, then Integrations, then Download WooCommerce plugin).
2. In your WordPress admin, go to Plugins, then Add New, then Upload Plugin. Click Choose File, select the .zip you just downloaded, then click Install Now.
3. Click Activate once the install finishes.
4. In the WordPress sidebar, open WooCommerce, then VritantAI.
5. Copy your API key from your VritantAI dashboard (Settings, then Integrations), paste it into the API key field, and click Save changes. The plugin verifies the key and subscribes the product webhooks for you.

A one-click install from the WordPress.org plugin directory is on the way and not yet available. Until that listing is published, the .zip upload above is the way to install the plugin.

Navigate to Data Hub in your VritantAI dashboard and click Sync Catalog. Product ingestion typically completes in 2 to 5 minutes for stores with fewer than 10,000 SKUs.

In Settings, then Widget, copy your embed script and paste it into your storefront theme just before the closing </body> tag. Your catalog-grounded shopping agent is now live for customers. Full instructions are in the Storefront Agent section below.

The shopping agent is grounded in your catalog. Make sure at least one catalog sync has completed successfully in Data Hub before embedding the widget.

Go to Settings, then Widget, and copy your unique embed script:

<script
  src="https://cdn.vritantai.com/widget.js"
  data-api-key="vc_your_api_key_here"
  data-theme="dark"
  async
></script>

Paste the script just before the closing </body> tag in your theme. The widget runs in a Shadow DOM and will not conflict with your existing CSS.

Go to Sandbox in your VritantAI dashboard to test the agent against your live catalog before your customers see it.

The WhatsApp agent is grounded in your product catalog. Confirm at least one catalog sync has completed in Data Hub before connecting.

Go to Settings, then Integrations, then WhatsApp and click Connect WhatsApp. A Meta popup will appear. Log in with the Facebook account that owns your WhatsApp Business Account.

Follow the steps in the popup to select your WhatsApp Business Account and phone number. VritantAI automatically discovers your account and subscribes the webhook. No manual configuration required.

After completing the popup, the settings page will show your connected phone number and WhatsApp Business Account ID. If you see these, the connection is live.

In Settings, then Widget, then WhatsApp, customise the opening message that greets new conversations. Make it specific to your store; mention your product category or a current promotion.

Use the Sandbox page to simulate a WhatsApp conversation from product enquiry through to checkout link generation. Verify that only real catalog SKUs appear in responses.

Conversation Intelligence clusters become meaningful once you have at least 50 agent sessions. New stores typically reach this within 1 to 2 weeks of launch.

Go to Analytics, then Conversation Intelligence, then VOC Clusters. Clusters are automatically labelled by intent theme (for example “size and fit”, “compatibility questions”, “price sensitivity”).

The SKU gap report surfaces queries that resulted in “no matching product” responses. These are your highest-value unmet demand signals. Consider adding these products or updating existing ones to match shopper language.

The shopper language library extracts the exact terms and phrases your customers use. Copy these into your PDP titles, bullet points, and meta descriptions to improve both SEO and AI citation relevance.

Go to Discovery, then Audits, paste a product page URL, and click Run Audit. Your first scorecard will be ready in under 60 seconds.

Your GEO score is a weighted composite: Schema Markup (50%), Page Performance (30%), Entity Coverage (20%). Scores above 80 indicate strong AI visibility.

Each audit surfaces a prioritised list of issues. Critical issues have the highest impact on citation rate and should be addressed first.

VritantAI auto-generates a patch with the correct schema. Here's an example of the Product schema it produces:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Brooks Ghost 16 Wide Running Shoe",
  "description": "DNA LOFT cushioning, engineered mesh upper, 2E width",
  "sku": "GHOST16-2E-BLK-105",
  "brand": {
    "@type": "Brand",
    "name": "Brooks"
  },
  "offers": {
    "@type": "Offer",
    "price": "114.95",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock"
  },
  "image": "https://example.com/ghost16-black.jpg"
}

Navigate to Discovery, then Audits, then [audit ID] and click Approve. VritantAI deploys the patch directly to your CMS via the Shopify GraphQL Admin API or WooCommerce REST API. Then run a benchmark under Discovery, then Citations: VritantAI queries Perplexity, ChatGPT, Gemini, Claude, and Copilot and reports citation rates per provider.

Go to Discovery, then Hallucination Monitor, then Configure. Set your preferred sweep frequency (daily or weekly) and the LLMs to query. All 5 providers are queried by default.

Add the product queries most likely to surface your brand. Examples: “best wide running shoes for men”, “[brand name] ingredient list”, “[product name] reviews”.

Each sweep generates hallucination events with severity (Critical, Warning, Info), exact extracted claims, and a side-by-side comparison to your catalog ground truth.

Mark events as Resolved once corrective action has been taken (for example, updated copy, patched schema). Dismiss low-severity events that are not actionable.