What is AI Knowledge Base by Macha? One search for every Zendesk knowledge source

AI Knowledge Base by Macha was a Zendesk Marketplace app (listing 1179926) that indexed your Help Center, macros, external websites and Shopify product pages, then recommended the right document for each ticket in the Zendesk sidebar. Paid accounts got up to 10 sources with up to 500 documents each. The app has been retired: multi-source knowledge search now lives in the Macha platform, priced by ticket volume from $299 a month for 750 tickets, where AI agents search your Help Center, websites, Notion, Confluence and Google Docs.

Question	AI Knowledge Base by Macha
Where it runs	Zendesk ticket sidebar (Zendesk Marketplace app)
Sources today	Zendesk Help Center, Zendesk macros, external websites, Shopify product pages
Listed as coming soon	Notion, Confluence
Status and price	Retired; the Macha platform starts at $299 a month for 750 tickets
Paid limits	Up to 10 sources, up to 500 documents per source
Free trial	On the Macha platform: $50 of usage, no credit card

If you're a Zendesk admin, you know the struggle. Your agents toggle between Help Center articles, internal Notion docs, Confluence pages, Shopify product information and website pages, all while trying to answer tickets quickly and accurately.

The cost of that is old but well documented: a 2012 McKinsey Global Institute report estimated that employees spend 1.8 hours every day, nearly a quarter of their workweek, searching for and gathering information. For support teams, scattered knowledge slows response times and shows up in customer satisfaction and agent morale.

AI Knowledge Base by Macha was our Zendesk app for this exact problem: it put every connected knowledge source into one AI search that lives inside the Zendesk ticket sidebar.

Why does scattered knowledge slow a support team down?

Most support teams have knowledge spread across several platforms. You might have:

• Help Center articles in Zendesk
• Internal processes documented in Notion or Confluence
• Product specifications on your website
• Inventory and product details in Shopify
• Policy documents in Google Drive
• Macros and canned responses in Zendesk

When a ticket comes in, agents pull from several of these. A SearchYourCloud survey found workers take up to 8 searches to find the right document and information, and that assumes they know where to look.

This fragmentation leads to:

• Slower first response times as agents hunt for the right information
• Inconsistent answers when agents can't find the current version of a policy
• More escalations when frontline agents can't reach internal documentation
• Agent burnout from constant context switching

What is AI Knowledge Base by Macha?

AI Knowledge Base by Macha is a Zendesk app that indexes your company's knowledge sources and makes them searchable from the ticket view, so agents don't leave Zendesk.

In practice it does four things:

• Suggests relevant documents for each ticket by reading the customer's message
• Searches across all sources using plain English queries
• Lives in the Zendesk sidebar, so agents never switch tabs
• Indexes content automatically, with no manual tagging

The app's listing names four supported sources today: Zendesk Help Center, Zendesk macros, external websites (scraped and indexed) and Shopify product pages. Notion and Confluence are listed as coming soon.

How does AI Knowledge Base work inside Zendesk?

Setup is short, and once configured the app runs in the background. Here's how it works:

Step 1: Connect your knowledge sources

Install the app from the Zendesk Marketplace, then open the Macha dashboard inside Zendesk or from the sidebar widget. Your Zendesk macros and Help Center are added automatically as the first two sources, with no configuration.

From there, you can add:

• External websites (the app scrapes and indexes the pages)
• Shopify product pages
• Notion pages (listed as coming soon)
• Confluence spaces (listed as coming soon)

Step 2: Automatic indexing and syncing

Once a source is connected, Macha fetches and indexes its content and keeps it in sync, so an updated Help Center article or a new website page shows up in results without anyone re-tagging it.

Zendesk's own documentation on knowledge management names keeping content current and relevant as one of the hardest parts of running a knowledge base. The app handles the indexing half of that problem for you; your team still owns whether the content itself is right.

Step 3: Recommendations for each ticket

When an agent opens a ticket, the app reads the customer's message and surfaces the most relevant documents from all connected sources, not only Help Center articles.

A "Recommended for this ticket" list might include:

• A Help Center article about returns
• A Shopify product page for the item the customer ordered
• A page from your website with your EU-specific refund policy
• A macro for handling this type of request

The agent doesn't have to search to see them.

Step 4: Natural language search

If the suggestions don't cover it, agents search every source in plain English. Type "exchange policy for EU orders" and results come back from the Help Center, macros, website pages and Shopify at once.

Agents don't need to remember which system holds which answer, or which keywords an article was written with.

How is it different from Zendesk's built-in Knowledge section?

1. External sources without a build

Zendesk default: The Knowledge section centers on your Help Center. Getting outside content in means setting up federated search, either a crawler for web pages or API work to push records.

AI Knowledge Base: Adds websites and Shopify product pages from the dashboard, with no API work and no plan upgrade on the Zendesk side.

2. No manual tagging

Zendesk default: Zendesk's best practices guide suggests you "add labels like troubleshoot and patch to the article to boost it in search results." That's ongoing maintenance.

AI Knowledge Base: Reads content for meaning, so new or updated documents become searchable without labels or tags.

3. One ranked list across sources

Zendesk default: Even with federated search, agents usually need a sense of which source is likely to hold the answer.

AI Knowledge Base: Reads the ticket and ranks results from every connected source together. An agent doesn't need to know whether the answer is a Help Center article, a macro or a Shopify page.

4. Plain-English queries

Zendesk default: As Zendesk's documentation explains, "article title length, keyword, or structure can influence instant search results," so agents adjust how they search to match how articles were written.

AI Knowledge Base: Accepts questions the way an agent would ask a colleague: "What's our return policy for damaged items?" It retrieves by intent, not by title wording.

5. No per-seat price

Zendesk default: The Knowledge section and federated search depend on your Suite plan and Knowledge tier, so access is tied to what you pay per seat.

AI Knowledge Base: The app is retired. Its successor, the Macha platform, bills per ticket rather than per seat, from $299 a month for 750 tickets, so the price is the same whether you have 5 agents or 500.

Zendesk's suggestions are helpful, but they draw on what's in, or pushed into, your help center. AI Knowledge Base applies the same idea to every source you connect.

What features does AI Knowledge Base include?

1. Multi-source connectivity

Search Zendesk Help Center, macros, external websites and Shopify product pages from one interface, with Notion and Confluence listed as coming soon.

2. AI recommendations

For every ticket, agents see the most relevant documents without searching. The app reads the customer's issue and pulls from all sources.

3. Natural language search

Agents search with conversational queries like "refund policy for EU orders" or "how to process warranty claims," with no search syntax to learn.

4. Native Zendesk integration

Everything lives in the Zendesk ticket sidebar. No tab switching and no separate tool.

5. Automatic syncing and indexing

When you update a document or add content to a connected source, the app re-indexes it, so agents see the current version.

6. Setup in minutes

No developer required. Install the app and it starts indexing your Zendesk macros and Help Center straight away. Add more sources from the dashboard whenever you like.

Which teams is AI Knowledge Base built for?

The app fits:

• Support teams with documentation spread across several platforms (Help Center, websites, product pages)
• Zendesk admins who want AI search without changing existing tools or workflows
• Growing teams who onboard new agents often and need them productive fast
• E-commerce teams with large catalogs whose agents need product specs from Shopify
• Teams supporting several languages whose documentation sits in separate sources

What changes for your agents day to day?

Here's what AI Knowledge Base changes in practice:

Faster first response times

When the relevant document appears with the ticket, agents read and reply instead of searching. That time comes straight off first response.

More consistent answers

With every source in one index, agents work from the same current information instead of whichever copy they found first.

Easier onboarding

New agents don't need to learn where each type of information lives. They search in plain English and get the answer.

Less tab switching

Everything happens in the Zendesk sidebar, so agents keep the ticket's context in front of them.

Fewer escalations

When frontline agents can reach deeper documentation, they resolve more tickets themselves instead of asking a senior teammate.

With recommendations on each ticket, the goal is that the right document is already on screen, and anything else is one plain-English search away.

How much does AI Knowledge Base by Macha cost?

Nothing now: the app has been retired and is no longer sold. Its knowledge search lives on in the main Macha platform, which has one plan priced by ticket volume, starting at $299 a month for 750 tickets, with setup and monitoring by the Macha team included. See the pricing page. The platform's free trial gives $50 of usage with no credit card.

How do you get started with AI Knowledge Base?

You can't install it any more, because the app has been retired. Start on the Macha platform instead: connect Zendesk, add your Help Center and any other knowledge sources, and build an agent that searches them on each ticket. The free trial covers that setup.

What sources are coming next?

None for the app, since it has been retired. The Macha platform already connects Notion, Confluence and Google Docs as knowledge sources. Questions about a source? Email [email protected].

Update (September 2026): Salesforce completed its acquisition of Fin, the AI agent company behind Intercom Fin, on 10 September 2026. Worth weighing in any long-term Intercom or Fin decision.

Common Questions About AI Knowledge Base

Is my data secure?

Yes. Macha processes data transiently and does not store ticket content long-term. Indexing and retrieval run over encrypted connections.

How long does indexing take?

It depends on the size of your sources, but most setups finish within minutes. Ongoing syncs run in the background.

Can I control which content is indexed?

Yes. You choose which sources to connect and can remove or update them at any time from the Macha dashboard.

What if I need a source that's not currently supported?

Email [email protected] with what you need, and we'll tell you whether it's on the roadmap. The main Macha platform already connects Notion, Confluence and Google Docs as knowledge sources for its AI agents, if you need those today.

Does this replace Zendesk's Help Center?

No. It works alongside it. Your Help Center stays your customer-facing documentation; the app makes it searchable together with your other sources.

Is AI Knowledge Base still worth installing?

It can't be installed now, because the app has been retired. If your agents regularly leave Zendesk to find an answer, the Macha platform covers the same need with AI agents that search every connected source, from $299 a month for 750 tickets. The McKinsey estimate puts the searching problem at 1.8 hours a day.

If your knowledge already lives in one well-tagged Help Center, Zendesk's native Knowledge section may be enough. If it's spread across macros, websites and Shopify, the app brings it into the sidebar where agents already work.