Build workflows
your agents run
Decide what your shopping agent does, from the first question to checkout. Then approve and monitor every run.
5/5 on G2
Trusted by 400+ brands and agencies
Build on a canvas
Chain triggers, agent steps and reviews without code.
Approve before it ships
Refunds, discounts and exceptions wait for your team.
Run and monitor
Every run, change and result in one timeline.
Everything in Agent Command Center
Visual workflows
Build on a canvas, no code needed.
Triggers
Start from a shopper question, a cart or a stock change.
Agent steps
Search, cart, checkout and support actions your agent runs.
Approvals
Nothing ships without your team.
Schedules
Daily, weekly or on demand.
Run history
Every run, change and result in one log.
Build, approve and run agent workflows
01
Build a workflow
Pick a trigger, add agent steps and a review.
02
Approve what your agent offers
Refunds and discounts wait for a yes.
03
Monitor every run
See what ran, what changed and what it moved.
Set up your agent in a few steps, then let it run
Tell Opttab what kind of business you run. It tailors the workflows, then runs them on schedule with your team in the loop.
Guided agent setup
Workflows tailored to your business
Approvals before anything ships
A closer look at Agent Command Center
CANVAS
Build on a canvas
Triggers, steps and approvals.
TRIGGERS
Start anything
Event triggers.
TEMPLATES
Start from a template
Ready-made workflows.
APPROVALS
Human in the loop
Review steps.
RUNS
Run history
Every run, logged.
MONITOR
Everything running, one view
Live workflow status.
ALERTS
Know when it fails
Failure alerts.
How Agent Command Center works for you
Build workflows on a canvas
Triggers, agent steps and approvals, without code.
Visual workflows
Triggers
Agent steps
Approve and monitor every workflow
Review offers, set limits and track every run.
Approvals
Schedules
Run history
What teams do with Agent Command Center
Answer sizing questions
Answer size and fit questions inside AI chats.
Cart recovery
Win back carts left in AI chats.
Order tracking
Answer “where is my order?” with live status.
Returns and exchanges
Start returns and swap sizes without a ticket.
Team alerts
Hand off to support when a shopper needs a person.
Client automation
Run the same workflows across clients.
FAQs
Can't find what you're looking for? Visit our Help Center.
What is the Opttab Brand Agent?
The Opttab Brand Agent is an AI agent that represents a brand inside AI assistants and on the brand's own website. It answers questions with live product data and can take the actions the brand allows, like building a cart or placing an order. Opttab hosts it as a remote MCP server for each brand and makes it available as a ChatGPT app, a Claude connector and an on-site shopping agent.
What is the Agent Command Center?
The Agent Command Center is where you build and manage your Opttab Brand Agent. You connect your data sources, choose which actions the agent can take, review what it's been doing, and switch actions or the whole agent on or off. It's the control panel for Opttab's Agentic Commerce pillar.
How do I set up a Brand Agent?
You set up an Opttab Brand Agent by connecting your data sources in Opttab: your product feed, authentication, CMS or commerce platform (such as Shopify) and shipping aggregators. Then you pick which actions the agent can take. Opttab creates a remote MCP server for your brand, for example at mcp.opttab.com/yourbrand. Opttab publishes it as a ChatGPT app and a Claude connector under its own developer account, and you can add it to your website as an on-site shopping agent.
Which systems does the Brand Agent connect to?
The Opttab Brand Agent connects to your product feed, authentication, CMS or commerce platform (such as Shopify) and shipping aggregators. It also connects to customer engagement platforms including Insider, Braze, Klaviyo and Salesforce Marketing Cloud, so journeys that start with AI carry on in your existing CRM and messaging flows.
Which actions can the Brand Agent take?
The Opttab Brand Agent can take the actions you turn on, such as product search, recommendations, cart creation, order placement, order status, returns, account actions and marketing sign-ups. You turn on each action yourself and can set limits on it. AI assistants can only call the actions you've turned on.
Who controls what the agent can do?
The brand controls what its Opttab Brand Agent can do. You choose which data the agent can use and which actions it can take, and you can switch off any action, or the whole agent, whenever you want. Every invocation gets a journey ID, so you can see each request, the action taken and the result.
What is a remote MCP server, and why does my brand need one?
A remote MCP server is an MCP server hosted online, so any compatible AI assistant can reach it over the internet without installing anything. Your brand needs one so assistants like ChatGPT and Claude can call your live catalog and actions instead of guessing from old web content. Opttab hosts one for each brand at mcp.opttab.com/yourbrand.
Do I need developers to build a Brand Agent?
No, you don't need custom development to create the agent itself, because Opttab builds and hosts the MCP server once you connect your sources and choose actions. Setup time depends on which tools you connect, and it can be done in hours. You do need access to your product feed, commerce platform and other systems to authorize the connections, and you should test the agent before it goes live.
Is the Opttab MCP for my data the same as the Brand Agent MCP server?
No, they're two different kinds of MCP access. API and MCP access to your Opttab data, from the Starter plan upward, lets your own AI tools read your visibility and analytics data. The Brand Agent MCP server, included on Enterprise for brands, and on Agency Scale and Agency Enterprise for agencies, is a public-facing server that lets AI assistants call your catalog and actions.
Which plans include the Brand Agent?
The Brand Agent with its own MCP server is included in Opttab's Enterprise plan for brands, along with dedicated onboarding. Agencies can offer Brand Agents to their clients through the agent module, which comes with the Agency Scale and Agency Enterprise plans. Book a demo to plan the setup for your catalog and markets.
How does Opttab keep the Brand Agent's information accurate?
The Opttab Brand Agent stays accurate by answering from the live data sources you connect, like your product feed and commerce platform, instead of from what an AI model remembers. So accuracy depends on keeping those sources up to date. The brand is responsible for correct prices, stock and shipping information, and Opttab may add confirmation steps for high-risk actions.
Do shoppers know they are talking to an AI agent?
Yes, shoppers know. The Opttab Brand Agent tells shoppers they're dealing with an AI system, in line with the transparency rules of the EU AI Act. Brands can't remove or hide this disclosure, and the agent can't be set up to claim it's human, wherever it runs.
Is the Brand Agent secure?
Yes, the Brand Agent is built with security controls. It runs through its own remote MCP endpoint with scoped authentication, and credentials for connected systems are stored encrypted in AWS Secrets Manager. You grant only the access the agent needs, every invocation is logged with a journey ID, and Opttab applies rate limits and abuse controls. You can turn off any action or the whole agent at any time.
Latest insights
Read the latest on AI visibility, GEO and agentic commerce.
Be the brand
agents choose.
Find, recommend, visit, buy. Opttab covers every step of the agentic journey.