"There's very few software applications that I myself as a CRO am like, 'I must use this every day' and Guru is one of them."
— Amanda Malko, Chief Revenue Officer, Thinkific

Company Overview

Thinkific is a leading online learning platform that empowers entrepreneurs and businesses – like Ametek, NASDAQ, Chargebee, Lululemon, Ironman – to create, market, and sell online courses and learning products.  With hundreds of employees supporting thousands of businesses globally, Thinkific runs a complex SaaS ecosystem: Salesforce for customer data, Gong for conversation intelligence, Zendesk for support tickets, Slack for internal communications, and Google Drive for documentation — among many other unstructured knowledge sources.

Like most fast-growing companies, Thinkific's knowledge lived everywhere.  And that meant their AI potential was limited by the quality, and trustworthiness, of what was behind it.

The Challenge: AI Is Only as Good as the Knowledge Behind It

Thinkific faced a compounding set of problems that many enterprises recognize: scattered knowledge, AI tools operating without a governed foundation, and a growing trust deficit in the accuracy of information.  None of these existed in isolation, they amplified each other.

‍

Problem 1: Knowledge Scattered Without Context

Daniel Andrew, Sr. Director of Finance and Operations, captures the cumulative cost:

"It's not that the information wasn't there or that it was not findable, just that typically you might have to look in a whole bunch of places and… when you start multiplying that time across the hundreds of people in our company and the number of times people need to make these requests… it actually becomes pretty meaningful."
— Daniel Andrew, Sr. Director of Finance and Operations

Individual searches weren't catastrophic — checking a Google Doc here, a Slack thread there, a Miro board somewhere else.  But the cumulative effect was staggering. Operations teams served as human search engines.  Requests like "What customers have been asking for this feature?" required combing through Gong transcripts, Zendesk tickets, Salesforce notes, and Slack, with no reliable, governed path to an answer.

‍

Problem 2: AI Tools Without a Governed Knowledge Foundation

Thinkific recognized AI's potential early.  Sales and customer success teams were already experimenting with AI tools for call prep and research.  But each operated in isolation, requiring manual data uploads, custom integrations, or copying from one system to another.  Teams were creating new fragmentation while trying to solve the original problem.

"We're inundated with data. It's not a lack of information. It's information overload."
— Amanda Malko, Chief Revenue Officer

Product and R&D teams needed synthesized customer insights to drive roadmap decisions, but accessing that intelligence meant manually aggregating Gong calls, Salesforce records, Zendesk tickets, and Slack conversations.  The signal was there.  A governed layer to surface it wasn't.

‍

Problem 3: Knowledge Decay Creating a Trust Deficit

The most insidious impact wasn't the time wasted, it was the decisions not made, or made on incomplete information.  Without a way to verify accuracy across distributed sources, teams couldn't trust insights for strategic decisions.  And without trust, every answer led to more questions about it’s accuracy and origins. 

The Approach: Guru as the Governed Knowledge Layer

What Thinkific needed wasn't another AI tool operating in a silo. They needed a governed knowledge layer, one that could connect their distributed sources, keep knowledge accurate automatically, enforce permissions, and power every AI tool and every team member from a single verified foundation.

That's why they chose Guru: the AI Source of Truth that structures and continuously strengthens company knowledge, governs it with citations and permission-aware access, and delivers trusted answers wherever people and AI tools work.

"Guru's positioning is really clear, and from where I sit, Guru adds value to our organization and we appreciate it."
— Amanda Malko, Chief Revenue Officer

‍

Connecting All Knowledge: Building the Governed Foundation

Guru connected to every system where Thinkific's operational knowledge actually lived — not just indexing content, but structuring and organizing it into a governed layer that preserved source permissions and kept knowledge accurate over time:

‍

Daniel noted the integration quality exceeded expectations:

"It's been pretty mind blowing watching, one, how easy it is to connect the sources, and then two, the quality of information we're getting in Guru from other places."

Amanda put it simply:

"Guru makes integration to other tools, the easiest I've ever seen across our systems."

‍

Governed Access by Design

Guru respected and enforced existing access controls across every source. Sales queried the customer data they had permission to see.  Finance accessed financial metrics. R&D accessed synthesized insights from Gong and Salesforce without needing direct licenses to those platforms.

The governed knowledge layer didn't create security gaps, it enforced the permissions already established in source systems, ensuring compliance and data integrity across every AI interaction.

The Solution: Deploying Knowledge Agents Purpose-Built for Every Workflow

With their governed knowledge foundation in place, Thinkific deployed Guru's Knowledge Agents — purpose-built agents that don't just answer questions, but structure, verify, and continuously improve the knowledge behind every answer. Each agent was configured for a specific workflow, pulling from the right sources and surfacing cited, permission-aware responses.

‍

The Call Prep Copilot: Customer Intelligence at Your Fingertips

The highest-impact deployment armed sales and customer success teams with instant, cited context before every customer conversation.

Daniel described it: "Our call prep copilot essentially arms our sales and customer success team with information from Salesforce, Gong, Zendesk, and several of the Slack channels where customer conversations might happen."

Amanda captured the shift: "It really is customer insights aggregated at your fingertips."

• Before: 30+ minutes of manual research across multiple systems before important calls
  
  
•  After: A single query — "What do I need to know for my call with XYZ?" — surfaces need-to-know context  with citations

Customer success feedback confirmed the impact: teams significantly reduced prep time by pulling account history, open tickets, past conversation themes, and CRM data into one cited, trusted response.

‍

Research Mode: Unlocking Previously Impossible Insights

Beyond answering direct questions, Guru's Research capability enabled sophisticated cross-source analysis that would have been prohibitively time-consuming to attempt manually.

Daniel shared one example: a research query identifying customers who had asked about stablecoins as a payment method, filtered to those above a specific GMV threshold. This required correlating data from the data warehouse, sales transcripts, and support tickets — none of it systematically tracked as structured data.

"I was really impressed with what came back in a research report that had a short list of customers, links to their support tickets where they had mentioned relevant topics. And then I was able to go and reach out to the customer success manager on the handful of customers that met the criteria that I was looking for."
— Daniel Andrew, Sr. Director of Finance and Operations

What would have required days of cross-functional coordination was completed in minutes — with full citations, so Daniel could verify findings and act immediately.

‍

The RFP Agent: Automating a High-Cost Workflow

Thinkific also deployed a Knowledge Agent to automate RFP responses — one of the highest-effort, lowest-leverage tasks in enterprise sales.

"We have an RFP agent, it writes our RFPs. Our sales team — it was the bane of their existence to do an RFP. Now we've automated the bulk of that."
— Amanda Malko, Chief Revenue Officer

Grounded in Guru's governed knowledge layer — product documentation, security policies, feature details, past approved responses — the agent produces accurate, on-brand, permission-aware drafts that reflect current, verified company knowledge. Sales time is redirected to higher-value activity.

‍

Results: From Fragmented to Governed

🕑 Minutes vs. Days - Research time savings

🚀 30+ min → Single query - Call prep time and effort reduction

📈 Bulk of process automated - RFP automation eliminating massive amounts of manual effort

‍

Across every deployment, teams experienced a consistent shift: from manually aggregating information across disconnected systems, to querying a single governed knowledge layer that surfaces cited, permission-aware answers in seconds.

A notable indicator of quality adoption: effective use required no prompt engineering expertise. The simplest questions — "Give me an update on this customer" or "What do I need to know for my call tomorrow?" — reliably surfaced the right information from multiple sources.

"Not needing the team to be really capable prompt engineers to still get value out of the tool, I think is a bit surprising, but in a really good way."
— Daniel Andrew, Sr. Director of Finance and Operations

‍

What's Next: Powering the Full AI Stack

Amanda sees the governed knowledge layer Thinkific has built as infrastructure. Not just for today's workflows, but for every AI tool and agent they'll deploy going forward.  

One near-term priority: using Guru to automate the flow of knowledge from engineering to customer-facing documentation, so product updates translate automatically into accurate, up-to-date help content, keeping the governed layer current without manual maintenance.

As enterprise AI moves toward agentic operations, the governed knowledge layer
becomes the critical differentiator.  Teams that have invested in structuring and verifying their knowledge will deploy agents that are safe, consistent, and auditable by design. Thinkific is already there.

‍