AI Co-Worker (MCP)

Ask your shop a question. Get the answer from your own records.

Connect a supported MCP assistant such as Claude to TaggIQ. It can look up customers, search connected supplier catalogues, draft quotes and list overdue invoices from your real data. It proposes, you approve.

Book an AI workflow assessment Start a 30-day trial

Want AI that works a job on its own? Meet the AI agents

  • Made for print, promo, sign and decoration businesses
  • Reads and writes TaggIQ records through MCP tools
  • Open MCP architecture

Co-Worker and agents

Two ways AI works in TaggIQ.

The AI Co-Worker answers when you ask and proposes work for you to approve. AI agents each run one workflow on their own, within your rules. Both work from the same TaggIQ records.

AI Co-Worker (this page)

You ask, it drafts, you approve.

Type a request in your MCP assistant: a quote, an overdue list, a production check. It reads and writes TaggIQ records through MCP tools and waits for your go-ahead on anything external.

AI agents

Each agent owns one job.

Following up unanswered quotes, finding customers due to reorder, chasing artwork and approvals, flagging thin margins and scheduling production. Start with one agent on your current tools or on TaggIQ.

Meet the AI agents

AI agents that help here

Before and after

One sentence instead of four supplier tabs.

Quoting a customer used to mean opening supplier tabs, copying SKUs, looking up their pricing tier and rewriting the same email again. Here is what that turns into.

The old way

  • Open several supplier tabs to find the right product
  • Copy SKUs into a spreadsheet
  • Look up the customer's last order for pricing
  • Calculate decoration costs by hand
  • Write the quote email from scratch
  • Hope you didn't typo a price

With the AI Co-Worker

You type

"Quote our corporate client for 100 navy polos with their logo, same tier as last time."

  • Claude looks up the customer and their pricing tier
  • Searches your connected supplier catalogues
  • Calculates decoration setup and per-unit costs
  • Drafts the quote with your branding
  • You review, approve and hit send

What you can ask

Your AI co-worker speaks promo.

The tools behind the chat read or write against your live TaggIQ data. Ask in plain English. The answer comes from what is actually in your shop.

  • Sourcing

    "Find me 200 USB drives under 5 each with one-colour print"

    Searches across your connected supplier catalogues.

  • Quoting

    "Quote our school client for 250 hoodies, embroidered front and back"

    Pulls customer history, applies their pricing tier, calculates decoration and builds the quote for your approval.

  • Customer intelligence

    "Show me everyone who hasn't ordered in 90 days"

    Segments the list, proposes the re-engagement campaign and waits for your approval before anything goes out.

  • Order ops

    "What's behind on production this week?"

    Status across connected orders, stages and supplier records, without leaving the assistant.

  • Cashflow

    "Which invoices are overdue, by how much, and for how long?"

    Prioritises the follow-ups, prepares reminders and waits for your send approval.

  • Reporting

    "How was March compared to February? Which product had the best margin?"

    Real numbers from your records, on demand. No spreadsheet needed.

  • Decoration pricing

    "What's the cheapest way to print 200 mugs in 3 colours?"

    Process suggestions, setup costs and tier breakdowns, all sourced from your own decoration rules.

  • Inventory

    "How many large navy polos do we have in stock?"

    Stock from connected supplier catalogues where feeds provide it, and your own warehouse.

  • Outreach

    "Draft a follow-up to anyone who got a quote 7+ days ago"

    Reads your tone, prepares the reply and never sends without your sign-off.

See it work

Three of the most common asks.

Quoting, sourcing across suppliers and chasing overdue invoices. Same chat, different tools. The demo loops through one example of each.

How it works

From signup to your first answer.

No engineer required. Connect a supported MCP-compatible assistant and it works from your TaggIQ records.

  1. Sign up and import

    Free 30-day trial. Bring your customers, suppliers and team. We help you migrate.

  2. Authorise your assistant

    Generate a Personal Access Token in Settings. Paste it into a supported MCP-compatible client such as Claude.

  3. Talk to your shop

    Ask for a quote, a status or an overdue list. It works from your products, customers and prices.

Claude

Supported MCP clients include Claude Desktop, Claude.ai and Claude Code. Other MCP-compatible clients may work depending on their MCP support. Generate one token in TaggIQ, paste it into your AI assistant, and keep control of what moves.

Trust by design

AI proposes. You decide.

Think of your AI co-worker as an operating layer over your TaggIQ records. It can search, calculate, build drafts and propose next actions. External actions stop until you approve them.

Your name is the only one on the email that gets sent. Every action is logged in your audit trail, and every approval is yours alone.

  • AI acts inside guardrails. You approve external moves.

    Quotes, orders and emails stay under your control. Nothing sends to customers, suppliers or books without your explicit confirmation.

  • Tenant-isolated by design

    Your data never leaks across organizations. AI sessions are bound to your org by Personal Access Token.

  • Money is read-only for AI

    AI cannot create payments, credit notes or change tax fields. Financial state is computed by TaggIQ alone.

  • External actions need a token

    Sending an email or creating a customer record requires a signed approval URL. Each one is single-use and expires in 24 hours.

  • Open standard, audited path

    Built on the open Model Context Protocol (MCP). Every tool call is logged and reviewable in your audit trail.

Related features

FAQ

AI Co-Worker questions.

What is MCP and why does it matter?

Model Context Protocol (MCP) is an open standard that lets AI assistants such as Claude securely connect to external tools and data, including your TaggIQ account. It means your AI can actually look up your customers, search your suppliers, and draft real quotes. Not just chat about them.

How is the AI Co-Worker different from TaggIQ AI agents?

The AI Co-Worker works when you ask: you type a request in your MCP assistant and approve what it proposes. AI agents each own one repetitive job, such as following up unanswered quotes or chasing artwork and approvals, and run it on their own within the rules you set, handing exceptions to a person. You can use either or both.

Which AI assistants work with TaggIQ?

Supported MCP clients include Claude Desktop, Claude.ai, and Claude Code. Other MCP-compatible clients may work depending on their MCP support. Generate one Personal Access Token in TaggIQ and connect it to the client you use.

Is my data safe when I use AI?

Yes. AI sessions are bound to your organization by your Personal Access Token. AI cannot directly create payments, send external emails, or cross tenant boundaries. Every action that affects external state requires your explicit approval. All tool calls are logged.

Do I need technical skills to set this up?

No. The setup is: copy a token from your TaggIQ Settings page, paste it into Claude's configuration. That's it. We have step-by-step guides for Claude Desktop and Claude.ai.

What happens if AI makes a mistake?

Mistakes are caught at the approval step. AI drafts a quote and you review it. AI proposes an email and you read it before it sends. AI suggests a customer reply and you approve before it goes out. Nothing reaches your customers, your books, or your suppliers without you in the loop.

How is this different from just chatting with a general AI tool about my business?

Generic chat tools can talk about your business in the abstract. A supported MCP assistant connected to TaggIQ works with the records and permissions inside your account: customers, products, pricing tiers, invoices, and the workflows around them.

Does this cost extra?

No. MCP integration is included on every TaggIQ plan, including the 30-day free trial. You only need a supported MCP-compatible AI client.

Start with the co-worker. Add agents when you are ready.

The AI Co-Worker is included on every plan, including the 30-day trial. Book an assessment to find the first job an AI agent should take on.

Book an AI workflow assessment Try the platform free for 30 days

Free assessment. No card for the platform trial. Agent fees are quoted separately.