Step inside Fable’s human behavior data lakehouse

We’re excited to introduce the Fable human behavior data lakehouse, a foundational part of our platform that will unlock powerful insights and enable human risk-based policy enforcement across our customers’ technology stack. 

This new architectural component underpins everything we do at Fable, and is designed to help security leaders answer the four questions that they seem to come back to again and again: 1. Who are our riskiest employees? 2. What material risk do they pose to the organization? 3. How do we reduce that risk? 4. How do we communicate all of this clearly to the board?

These are deceptively hard questions. Answering them requires stitching together data from across the enterprise—HR, IT, and security systems—to build a clear and contextual picture of each employee: who they are, what they have access to, how they behave over time, and what business functions they support. Historically, this kind of integration has been nearly impossible because this information lives in silos and wasn’t built to work together.

That’s why we took a fundamentally different architectural approach. Instead of pre-defining a rigid data schema and building flexibility at the application layer, we flipped the script. Modeled after the industry-standard Medallion architecture, our data pipeline organizes human risk data into bronze, silver, and gold layers—ensuring we capture everything, make it consistent and connected, and deliver decision-ready insights. This structured approach means faster analysis, higher data quality, and a stronger foundation for advanced analytics, machine learning, and real-time interventions. It allows us to identify concrete behavior signals for every individual, uncover relationships and patterns between data points, calculate risk dynamically, and deliver interventions in the moment.

This is what sets Fable apart. Most human risk solutions can tell you who your riskiest employees are—and possibly why. And some of the smarter ones may pull a few pre-selected attributes to display in your dashboard or report. But they stop short of incorporating deeper business context, dynamically updating risk, and turning that intelligence into real-time decisions. With our data lakehouse, we don’t just let you report on risk—we help you interrogate it, understand it, explain it, and reduce it. 

Here are some examples of questions you can ask because of this platform choice: 

• Are there elevated risks for employees traveling for work in China? What access do they have to our sensitive IP, and what should I do about it?
• Are there any managers with former employees who still have access to crown jewel apps? Share the list and show them how to deprovision.
• Which offices have the lowest rate of password manager usage? Which have a high number of employees with compromised credentials?
• Which employees with access to customer information have a high number of data-handling alerts? Do any have a departure date already in our HR system?
• Who is at risk for the latest Scattered Spider attack? Show me the employees at my organization who need to be prepared.

Another powerful aspect of our data lakehouse is how it pairs with generative AI. This combination will enable customers to do two key things: 1. Automatically convert paper-based policies into machine-readable formats, allowing for detection and analysis of out-of-policy behavior across the organization; and 2. Query Fable efficiently with a natural-language conversational agent. Lots of products use AI to enable natural-language queries, but with varying results. The good news is our data lakehouse retains the raw data, so every analysis is grounded in evidence you can verify. That means more accurate, explainable results, and guardrails against LLM hallucinations that can plague systems without a traceable source of truth.

Looking ahead, we’re evolving toward an agent-based experience—where you can go beyond simply asking questions and performing analysis, to taking action. The next step for us will be to extend the definition of “interventions” in Fable to include agentic workflows that execute policy changes in third-party solutions. This means turning human risk that you learn about in Fable into action across your technology stack. The Fable human behavior data lakehouse is the backbone that makes it all possible.