How Do You Connect Knowledge Sources to AI on Freshdesk? (2026)

To connect knowledge sources to AI on Freshdesk, open AI Agent Studio, select your Freddy AI Agent and go to Build → Knowledge, where you can add solution articles, public URLs (10 per agent), uploaded PDF, DOCX or TXT files (up to 35 MB each), custom Q&As and app sources such as Google Drive and SharePoint. An AI agent layer like Macha adds website crawls, uploaded spreadsheets, and live reads of Google Docs, Notion and Confluence on top. The rest is the part that decides accuracy: cleaning the content and measuring what the agent resolves. (New to Freshdesk's KB itself? Start with Freshdesk knowledge base explained.)

Why do knowledge sources decide how well Freshdesk AI answers?

Modern support AI runs on a pattern called retrieval-augmented generation (RAG). Instead of answering from whatever a language model memorized during training, the agent first retrieves the most relevant passages from your connected content, then writes an answer grounded in them. The model supplies the language; your knowledge supplies the truth.

That design has one blunt consequence: the agent can only resolve what it can find. Three things follow directly from the quality of your sources:

• Accuracy: answers come from your documented policies and steps, not a plausible-sounding guess.
• Fewer hallucinations: when the answer exists in your content, the agent uses it; when it doesn't, a well-built agent escalates instead of inventing.
• Freshness: update an article and the agent's answers move with it, as long as that source re-syncs or is read live.

So before you tune a single prompt, the most useful move is getting the right sources connected and clean. Everything below serves that.

Which sources can Freshdesk's native Freddy AI Agent read?

Freshdesk's built-in AI lives in AI Agent Studio. You configure what it learns from under AI Agent Studio → [your AI Agent] → Build → Knowledge (Freshdesk docs). The Freddy AI Agent learns from five source types (Freshdesk docs):

• Solution articles: your native Freshdesk knowledge base (the "Solutions" content store behind your Help Center).
• URLs: help centers, product docs, policy pages, or FAQ sections on the open web. Limits are 10 URLs per AI Agent and 25 per account, with up to 3,000 pages per site. Private URLs can't be learned.
• Files: PDF, DOCX, and TXT documents, up to 200 files and 35 MB per file.
• Custom Q&As: hand-written answers for questions your articles don't cover yet. Freshworks sets no strict limit but recommends keeping them concise.
• Apps: Google Drive and SharePoint, up to 100 shared drives or sites per AI Agent, with a 50 MB maximum file size.

Freshworks sets those limits deliberately, to keep the agent focused on high-quality content rather than the entire internet. That's a feature, not a bug: a smaller, curated source set almost always beats a sprawling one.

The native path is a solid place to start, especially if your knowledge already lives in Freshdesk Solutions. Two things to budget for: Freddy AI Agent usage is billed in sessions, with 500 sessions given once per account and then $49 per 100 sessions (Freshdesk pricing). And where teams hit limits is breadth of source types and control over retrieval and escalation, which is where an AI agent layer comes in.

Which sources can an AI agent layer like Macha connect on Freshdesk?

Macha isn't a help desk and it isn't a knowledge base. It's an AI agent layer that runs on top of the help desk you already use (Zendesk, Freshdesk, Gorgias, Front, HubSpot or Intercom), reading from the knowledge sources you connect and resolving issues inside the same ticket, handing off to a human with full context when it isn't confident. Because it sits as a layer rather than a feature inside one product, it can blend a wider mix of sources into a single agent:

• Freshdesk solution articles: the Freshdesk connector includes a Search Articles tool, so the agent can query your existing KB by keyword.
• Any public web page: add a single page as a source.
• Website crawl: pull in up to 200 pages from a docs site or help center in one source.
• Uploaded files: PDF, DOCX, TXT, CSV, and XLSX, up to 10 files at a time and 20 MB each.
• Google Docs, Notion and Confluence: connected as live sources, so the agent reads the current version of an internal wiki page or doc every time rather than a copy.
• Past-ticket analysis: results from bulk ticket studies can be pushed in as a searchable source, turning "how we've answered this before" into knowledge.

The practical win is that one agent can read your public help center and your internal docs at once, not just a single silo. For a deeper, platform-neutral walkthrough of the same idea, see how to connect your knowledge base to an AI agent.

How does ingestion work: crawl, chunk, embed, retrieve?

You don't have to manage any of the plumbing, but understanding it tells you why good structure matters so much. When you connect a static source (a file or a crawled site), four things happen:

1. Crawl or upload. The platform fetches your content: crawling a docs site page by page, or reading the files you upload.
2. Chunk. Each document is split into smaller passages so retrieval can return a focused, relevant slice rather than a whole 3,000-word page. In Macha, spreadsheets are chunked by row, about 75 rows per chunk with the header row kept, and small spreadsheets of 2,000 rows or fewer skip chunking and go straight into the agent's context.
3. Embed and index. Each chunk is turned into an embedding (a mathematical representation of its meaning) and stored in a vector index for semantic search.
4. Retrieve and answer. When a customer asks something, the agent searches for the best passages, then writes a grounded answer, and can pull a full document on demand when it needs the detail.

Live sources skip all four steps. For Google Docs and Notion, Macha attaches read tools (google_read_doc, notion_get_page) and the agent fetches the current page each time it needs it.

Freshness is the part teams forget. Static copies don't update themselves: in Macha an uploaded file has to be deleted and re-uploaded when the original changes, and a crawled website needs a Resync. If your agent is quoting a price you changed last quarter, the culprit is almost always a stale source, not a "dumb" model. Content that changes weekly belongs in a live source.

One cost note worth knowing: Macha doesn't bill indexing or storage separately. You're charged per ticket: one conversation, charged once, however many articles the agent reads and however many replies it takes. So connecting a large knowledge base adds nothing to the bill, and a well-grounded answer that took six lookups costs exactly the same as a lucky one-liner. Freddy's model is different: every AI Agent session past the one-time 500 costs money, whatever the answer quality.

How do you connect your knowledge sources to AI on Freshdesk, step by step?

The same shape applies whether you're configuring native Freddy or an agent layer:

1. Audit what you have. Pull your top 15–20 ticket reasons from Freshdesk reporting. That list is your priority source list: connect the content that answers those first.
2. Connect your Freshdesk Solutions KB. This is your highest-signal source: it's already written for customers and already maps to real questions.
3. Add the gaps as extra sources. Docs site, policy pages, internal Notion/Confluence wikis, key PDFs, and a handful of custom Q&As for questions no article covers yet.
4. Let it index, then wait for the "ready" signal. Ingestion isn't instant. In Macha, uploads index in seconds to a couple of minutes and a crawled site in a few minutes; the source shows Ready when done. Don't test before it's finished.
5. Scope each source to the right agent. A refund agent doesn't need your engineering wiki. Give each agent only the sources relevant to its job; tighter scope means cleaner retrieval.
6. Test with questions you know the answers to. Ask 20–30 real customer phrasings and confirm the agent uses the right passage, not a vaguely related one.
7. Set the escalation path. Decide what happens when the agent can't find an answer: hand off to a human with context, never guess.
8. Schedule re-syncs. Resync crawled sites and replace uploaded files whenever they change, or move fast-changing content to a live source.

How should you structure and clean your KB so AI answers well?

This is where most of the resolution rate is won or lost. Retrieval rewards content that's structured the way a searcher thinks:

• One job per article. A focused article ("Why is my order stuck on processing?") chunks cleanly and retrieves precisely. A 4,000-word mega-page buries the relevant passage among nine irrelevant ones.
• Write titles in the customer's words. "Reset your password" beats "Authentication credential management." Retrieval and customers both search in plain language.
• Put the answer near the top. Lead with the resolution, then add detail. Front-loaded passages retrieve and read better.
• Kill duplicates and contradictions. Three conflicting versions of your refund policy will produce three different agent answers. Keep one source of truth per topic.
• Retire stale content. The agent faithfully repeats whatever you leave connected. Old screenshots and outdated steps are worse than nothing.
• Curate, don't dump. Connecting everything lowers answer quality. Connect your good content and add deliberately.

If you're building or rebuilding the underlying Freshdesk KB itself, our Freshdesk knowledge base explained guide covers the category → folder → article structure, Help Center, and plan gating in full.

How do you turn failed searches into articles to write?

The single most valuable report you'll run isn't "what did the AI answer". It's "what did it fail to answer." Every question the agent couldn't resolve, and every customer search that returned nothing, is a precise instruction for what to write next. Run the loop continuously:

1. Capture misses. Pull the agent's low-confidence and escalated conversations, plus zero-result help-center searches.
2. Cluster them. Group the misses into themes; you'll usually find a handful of topics driving most of the gaps.
3. Write or fix the source. Add the missing article (or a custom Q&A as a fast patch), or correct the one that retrieved badly.
4. Re-sync and re-test. Confirm the next batch of those questions now resolves.

Do this for a few cycles and your resolution rate climbs without touching the model at all, because you're feeding the part that actually decides answers: the knowledge.

How do you measure deflection on Freshdesk?

"Deflection" gets thrown around loosely, so define it before you report it. Three layers, smallest to largest:

• Deflection: the customer found an answer (in self-service or via the agent) and didn't open a human ticket. This is a containment metric.
• Resolution: the agent actually solved the issue, end to end, in the conversation. Stronger than deflection; it means the problem went away, not just that the ticket didn't get created.
• Automation: the agent did real work (looked something up, took an action) whether or not it fully closed the case.

To measure it on Freshdesk, watch: deflection/containment rate (share of conversations resolved without a human), AI resolution rate (fully solved by the agent), escalation rate and reasons (your gaps list), CSAT on AI-handled conversations (resolved isn't the same as happy), and answer accuracy spot-checks (sample AI replies and grade them). Be honest about the denominator: a 60% deflection rate on only the easy half of your tickets is a different claim than 60% across everything.

A fair expectation: AI deflection scales with KB coverage. Repetitive, well-documented question types deflect heavily; novel or account-specific issues won't, and shouldn't. The number you can defend is the one tied to how much of your real ticket mix your sources actually cover.

Can AI fix a weak knowledge base?

No. An AI agent is only as good as the knowledge you connect it to. Garbage in, confident-wrong out. No model, native or layered, fixes a thin or contradictory knowledge base; it just repeats it faster. Everything on this page (clean structure, current articles, customer-worded titles, gaps filled from real misses) is what makes both Freshdesk's native Freddy and an AI agent layer actually resolve tickets. Get the sources right first; the AI rides on top of them.

If your ticket mix is mostly repetitive questions your help center could answer, an agent layer that reads your Freshdesk KB plus your wider docs is a fast way to test how far that takes you. You can start with $50 of free usage, no credit card, and connect your sources without touching your Freshdesk setup. After the trial, Macha is one plan priced by ticket volume, from $299 a month for up to 750 tickets, about $0.40 a ticket. You don't have to wire up the sources yourself: on the trial and every plan, the Macha team builds the knowledge base from your help center, website, documents and past tickets, fills the gaps for the first category, builds the agent and runs it in safe mode until the drafts are right. The price also covers a dedicated success manager who handles your changes afterwards, such as a new source, an instruction change or a new tool. (For the full picture of running Macha on Freshdesk, see Macha for Freshdesk.)

Frequently asked questions

What is a Freshdesk AI knowledge base? It's the set of knowledge sources an AI agent reads to answer customer questions on Freshdesk. Natively, Freshdesk's Freddy AI Agent learns from your solution articles, uploaded files, public URLs, custom Q&A pairs and connected Google Drive or SharePoint, configured in AI Agent Studio under Build → Knowledge. An AI agent layer like Macha can search the same Freshdesk KB plus crawled websites, uploaded spreadsheets, and live Google Docs, Notion and Confluence pages.

Which knowledge sources can I connect to AI on Freshdesk? With native Freddy: solution articles, PDF/DOCX/TXT files (200 files, 35 MB each), public URLs (10 per agent, 25 per account), custom Q&As, and Google Drive or SharePoint. With Macha: your Freshdesk articles through the connector, any public page or a 200-page website crawl, uploaded files (PDF/DOCX/TXT/CSV/XLSX, 20 MB each), and live Google Docs, Notion and Confluence.

How do I train AI on my Freshdesk help center? Connect your Solutions/help-center content as a knowledge source, let it index, scope it to the right agent, then test with real customer phrasings. The agent retrieves the most relevant passages and answers from them: you're not "training" a model so much as connecting and curating the content it retrieves from.

How long does ingestion take? It's not instant. Freshworks' current docs don't give a time for Freddy. In Macha, uploaded files index in seconds to a couple of minutes and a crawled website usually in a few minutes. Wait for the Ready status before testing.

Does connecting a big knowledge base cost more? Not with Macha: indexing and storage aren't billed separately. Billing is per ticket: one conversation, charged once, from $299 a month for 750 tickets (about $0.40 each), and that includes the Macha team building the knowledge base for you and a dedicated success manager who handles your changes afterwards. Always confirm current details on the pricing page, and start with a free trial ($50 of usage, no credit card required). With Freddy, cost follows AI Agent sessions rather than source size.

Will the AI make things up? A well-built agent answers from your connected sources and escalates to a human when the answer isn't there. Grounding in good knowledge plus a clear escalation path is what keeps it honest, which is exactly why source quality matters more than model choice.

How do I measure deflection on Freshdesk? Track deflection/containment rate, AI resolution rate, escalation rate and reasons, CSAT on AI-handled conversations, and periodic accuracy spot-checks. Be honest about the denominator: deflection on the easy subset of tickets isn't the same as deflection across your whole volume.

How much does the Freddy AI Agent cost to run? Freshdesk gives 500 Freddy AI Agent sessions once per account, then charges $49 per 100 sessions. Check the Freshdesk pricing page for your plan.

What should you connect first?

Connect your Freshdesk Solutions KB first, then fill the gaps with docs, files, and internal wikis, structure articles for retrieval, run the failed-search-to-new-article loop, and measure deflection against your whole ticket mix. Whether you use native Freddy or add an AI agent layer on top, your knowledge sets the ceiling on what AI can deflect. Let the misses tell you what to write next.

Checked against Freshworks' Freshdesk documentation and Macha's docs on 24 September 2026. Freshworks revises AI Agent Studio features, source limits, and packaging periodically; confirm current details on freshworks.com before relying on them.