NEW Autonomous agents for online commerce

AI agents that run your store around the clock.

Mac Quan Shop deploys specialized AI agents that answer customers, recommend products, manage inventory and run campaigns, so your team can focus on growth.

No-code setup Human approval on key actions Full activity logs
agent.run: support-agent
[10:42:03] trigger new ticket #4821 "Where is my order?"
[10:42:04] lookup order 88134 -> status: shipped
[10:42:05] tool carrier.track(88134)
[10:42:06] reply "Your order left our warehouse today..."
[10:42:06] resolved no escalation needed
[10:43:10] trigger inventory low: SKU-204 (3 left)
[10:43:11] action draft restock PO, awaiting approval
24/7
Always-on coverage
Multi-agent
Agents that share context
Human-in-the-loop
Approve high-impact actions
Audit trail
Every decision is logged
The agents

Specialized agents for every part of your store

Each agent has a clear role, a defined set of tools and guardrails you control.

01

Support Agent

Answers tickets and chat messages instantly, looks up orders and escalates edge cases to your team.

ChatEmailOrder lookup
02

Sales Agent

Recommends products based on intent, handles cart recovery and upsells at the right moment.

RecommendationsCart recovery
03

Inventory Agent

Forecasts demand, flags low stock and drafts purchase orders for your approval before anything is sent.

ForecastingRestock alerts
04

Marketing Agent

Plans and tests campaigns, writes copy variants and reports which messages convert.

CampaignsA/B testingReports
How it works

From connection to autonomy in three steps

Most stores are live within a day. No engineering team required.

01

Connect your stack

Link your storefront, helpdesk, inventory and email tools with secure, read-scoped access you can revoke at any time.

02

Set goals and guardrails

Define what each agent can do alone, what needs approval and what it must never do. Tone, limits and budgets are yours to set.

03

Monitor and improve

Review the activity log, approve actions in one click and give feedback that makes every agent sharper over time.

Capabilities

Built for real operations, not demos

Agents use tools, remember context and hand off cleanly to people when a situation needs judgment.

  • Tool useAgents call your APIs to check orders, update tickets and read stock levels.
  • Shared memoryCustomer history and product context are shared across agents, so nothing is repeated.
  • GuardrailsSpending limits, refund thresholds and blocked actions are enforced before execution.
  • Clean handoffWhen an agent escalates, your team gets the full conversation and a suggested next step.
agents/inventory.yaml
agent: inventory-agent
goal: keep SKU stock above safety level
tools: [inventory.read, po.draft]
autonomy:
  auto: [forecast, alert]
  approval: [po.send]
  blocked: [price.change]
budget: max_po_value: 5000 USD

Put your first agent to work this week

Tell us about your store and we will set up a pilot agent with you, free of commitment.

Contact

Talk to the team

Share your platform, catalog size and biggest bottleneck. We reply within one business day.