Definitions, product areas, and practical use cases for every key concept in the One Model platform and One AI. Click any term to expand its full definition, see where it appears in the product, and find related terms.
Last updated: February 2026 | 60+ terms | Est. reading time: 6 min
On This Page
Platform Terms
Allocation · Alternative Calendar · Annotations · Average Headcount · Branding · Change Indicator · Chart Color Palette · Column (Drill Through) · Contextual Security · Cumulative · Data Destination · Data Mesh · Data Snapshot · Data Source · Data Stories · Deep Linking · Dimension · Drillthrough · Entity Relationship Diagram · Explore · Geomap · Insight Library · List Report · Metric · Multi-Page Storyboard · Org Chart · People Data Cloud · Pin (Pinning) · Pivoted Dimension · Point in Time · Proxy · Query · Role-Based Security · Role Sync · Site Validation · Sparkline · SQL Explorer · Storyboard · Storyboard Library · Tile · Time Function · True Blanks · Waterfall Chart · Zero Results / Zero Values
One AI & Machine Learning Terms
Augmentation · Classification Model · Cross Validation · Deploy · EDA Report · Embedded Insights · Feature Engineering · Forecasting · Global Settings · One AI · One AI Assistant · Persist · Recipe · Refinement · Regression Model · Results Summary Report · Risk Model · SHAP Values · Table Insights · Upsampling
One Model Platform Terms
-
A data modeling technique that distributes or assigns values (such as cost or headcount) across organizational dimensions when source data doesn't provide a direct mapping. Allocations let you attribute shared resources or costs to specific business units, departments, or cost centers within One Model's data model.
📍 Product area
People Data Cloud → Data Modeling & Transformation
⚡ Use case
Your finance team reports salary costs at a division level, but your analytics team needs cost-per-department breakdowns. You configure an allocation rule to proportionally distribute division-level costs across departments based on headcount ratios.
Related: Dimension · Metric · True Blanks
-
A calendar configuration that overrides the default Gregorian calendar to align One Model's time periods with your organization's fiscal year, pay periods, or custom reporting cadence. Alternative calendars ensure that metrics like "Q1 Headcount" reflect your company's definition of Q1, not the standard January–March.
📍 Product area
People Data Cloud → Data Modeling & Transformation
⚡ Use case
Your company's fiscal year runs April to March. You configure an alternative calendar so that when a leader filters a Storyboard to "FY2026 Q1," they see data from April–June, not January–March.
Related: Time Function · Point in Time · Cumulative
-
Textual elements added to data visualizations within Storyboards to provide additional context, highlight specific insights, or clarify scope. Annotations can include labels, callouts, emoji symbols, or explanatory notes positioned on or near chart elements.
📍 Product area
Using Storyboards → Charts & Visualizations
⚡ Use case
A waterfall chart shows an unexpected spike in voluntary terminations in Q3. You add an annotation reading "Restructure announced Aug 15" directly on the chart so stakeholders immediately understand the context.
Related: Storyboard · Tile · Waterfall Chart
-
A calculated metric that averages the number of active employees across a span of time, rather than counting them at a single point. Average headcount is the standard denominator for rate-based metrics (Termination Rate, Hire Rate, Promotion Rate) because it prevents overstating or understating results that can occur with point-in-time counts.
📍 Product area
Using Explore → Metrics & Time Concepts
⚡ Use case
You're building a Termination Rate metric. Using average headcount as the denominator gives a more accurate rate because it accounts for the population size across the entire period, not just the final day.
Related: Metric · Cumulative · Point in Time
-
Platform configuration settings that allow administrators to customize the visual appearance of their One Model instance with organization-specific colors, logos, and styling. Branding applies to the navigation, Storyboards, login screen, and exported content.
📍 Product area
Admin Toolkit → Platform Configuration
⚡ Use case
Before rolling out One Model to 200 HRBPs, your admin applies corporate branding: logo, primary colors, and fonts, so the platform feels like an internal tool, not a third-party application.
Related: Storyboard · Chart Color Palette
-
A visual element on a Storyboard tile that shows the direction and magnitude of change in a metric compared to a prior period. Change indicators typically display an arrow (up/down) and a percentage or absolute value, giving viewers an instant sense of trend without reading the full chart.
📍 Product area
Using Storyboards → Charts & Visualizations
⚡ Use case
On an executive Storyboard, the "Voluntary Turnover" tile shows the current rate alongside a red upward arrow and "+1.2pp vs. prior quarter," immediately flagging that turnover has worsened.
Related: Sparkline · Tile · Storyboard
-
A configurable set of colors applied to charts and visualizations across Storyboards. The palette ensures visual consistency and can be customized to match organizational branding or accessibility requirements.
📍 Product area
Using Storyboards → Design & Layout
⚡ Use case
Your brand guidelines specify a particular set of colors for data visualization. You update the chart color palette so every chart produced by any user automatically uses the approved colors.
Related: Branding · Storyboard · Tile
-
A data field configured to appear when a user drills through from an aggregate metric to see individual-level detail. Drill through columns are set up by admins to control which personal or record-level fields (e.g., name, employee ID, department) are visible, and access is governed by role-based security.
📍 Product area
Admin Toolkit → Admin Tools & Platform Features
⚡ Use case
When a manager drills through on a "Headcount by Department" metric, they see columns for Employee Name, Hire Date, and Job Title, but not Salary, because their role doesn't have access to compensation columns.
Related: Drillthrough · Role-Based Security · List Report
-
A dynamic security model that grants data access based on the user's relationship to the data, such as their position in the organizational hierarchy. Unlike static security (where access is manually assigned to fixed groups), contextual security automatically adjusts as organizational structures change over time.
📍 Product area
Security & SSO → Authorization & Role-Based Security
⚡ Use case
A VP of Engineering can see people data for everyone who reports up to them. When a restructure moves a team from Engineering to Product, the VP's access automatically updates with no admin intervention needed.
Related: Role-Based Security · Dimension · Proxy
-
A time concept applied to metrics that sums data across a span of time, accumulating values from the start to the end of the selected period. Cumulative metrics are used for event-based counts like total hires, total terminations, or total transfers within a year.
📍 Product area
Using Explore → Metrics & Time Concepts
⚡ Use case
You create a "Cumulative Hires" metric for 2026. When a user selects the full year, they see the running total of all hires from January through December, not just a single month's count.
Related: Point in Time · Average Headcount · Metric
-
A configured output endpoint that sends processed data from One Model to an external system such as Snowflake, Azure, an SFTP server, or another analytics platform. Data destinations allow you to push One Model's modeled, analytics-ready data outward for consumption by other tools.
📍 Product area
People Data Cloud → Data Destinations
⚡ Use case
Your BI team uses Tableau for executive dashboards. You configure a data destination to push curated workforce data from One Model into your Snowflake data warehouse each night, where Tableau queries it.
Related: Data Source · Data Mesh · People Data Cloud
-
One Model's data ingestion and integration layer (also referred to as People Data Cloud). Data Mesh automates the extraction, cleansing, and modeling of HR and workforce data from multiple source systems into a unified, analytics-ready data model. It connects to HRIS platforms (Workday, SuccessFactors, Oracle), recruiting systems (Greenhouse), surveys, and custom file sources.
📍 Product area
People Data Cloud
⚡ Use case
Your organization uses Workday for core HR, Greenhouse for recruiting, and Qualtrics for engagement surveys. Data Mesh ingests data from all three systems nightly, resolves entity conflicts, and presents a single joined data model.
Related: Data Source · Data Destination · Entity Relationship Diagram · People Data Cloud
-
A point-in-time capture of your data model that preserves the state of your data at a specific date. Snapshots enable historical comparison and auditing, allowing users to see what the data looked like at any previous processing date.
📍 Product area
People Data Cloud → Advanced Data Scenarios
⚡ Use case
During an audit, you need to verify what headcount data looked like on December 31, 2025. You access the data snapshot from that date to confirm the numbers reported at year-end.
Related: Data Mesh · People Data Cloud
-
Any system, file, or connection from which One Model ingests data. Data sources can be API-based connectors (Workday, Greenhouse, SuccessFactors), file-based (CSV, Excel flat files via SFTP or direct upload), or custom integrations. Each data source is configured with extraction logic, scheduling, and field mapping.
📍 Product area
People Data Cloud → Data Acquisition
⚡ Use case
You're onboarding a new survey platform. You configure it as an external SFTP data source in One Model, map the survey response fields to the data model, and schedule nightly extraction.
Related: Data Mesh · Data Destination · People Data Cloud
-
One Model's product name for its interactive data visualization and storytelling capability, delivered through Storyboards. Data Stories is the marketing umbrella for the Storyboard feature set, emphasizing the narrative and communication aspect of analytics.
📍 Product area
Using Storyboards (overall product feature)
⚡ Use case
A People Analytics leader uses Data Stories to build a multi-page Storyboard that walks the CHRO through the complete talent acquisition pipeline narrative, from open requisitions to offer acceptance rates.
Related: Storyboard · Multi-Page Storyboard · Tile
-
A feature that creates direct URL links from within One Model Storyboards to external systems such as your HRIS (e.g., Workday, SuccessFactors). Deep links allow users to click on a person or record in a list report and jump directly to that individual's profile in the source system.
📍 Product area
Admin Toolkit → Admin Tools & Platform Features
⚡ Use case
An HRBP views a list report of employees with upcoming work anniversaries. They click on an employee's name, and the deep link takes them directly to that person's Workday profile to initiate a recognition action.
Related: List Report · Storyboard · Drillthrough
-
A categorical attribute used to slice, filter, and group metric data in Explore and Storyboards. Dimensions represent organizational hierarchies (department, business unit, location), employee attributes (gender, tenure band, job family), or time-based categories. Dimensions can be multi-level (e.g., Region → Country → City) and are governed by role-based security.
📍 Product area
Using Explore → Explore Basics; Security & SSO → Authorization
⚡ Use case
You run a headcount query in Explore and add "Location" as a dimension to see headcount broken down by country. You then pivot by "Gender" to see the gender distribution within each country.
Related: Pivoted Dimension · Metric · Role-Based Security
-
The ability to click on an aggregate metric value in a Storyboard or Explore query and see the individual records that make up that number. Drillthrough reveals person-level data (governed by security permissions) behind an aggregate, enabling verification and operational action.
📍 Product area
Using Explore; Using Storyboards
⚡ Use case
A Storyboard shows "42 Voluntary Terminations in Q3." The HR Director clicks on "42" and drills through to see the list of individuals, their departments, tenure, and termination reasons.
Related: Column (Drill Through) · Role-Based Security · List Report
-
A visual representation of the data tables in your One Model instance and the relationships between them. The ERD shows how employee records, event tables (hires, terminations, transfers), dimension tables (departments, locations), and fact tables connect. It is auto-generated from your specific data model configuration.
📍 Product area
People Data Cloud → Data Model & Architecture
⚡ Use case
A data engineer new to your One Model instance reviews the ERD to understand how the employee table joins to the compensation table and which dimension tables are available for reporting.
Related: Data Mesh · People Data Cloud · Dimension
-
One Model's primary data query and analysis tool. Explore allows users to build ad hoc queries by selecting metrics, dimensions, time periods, and filters, then visualize results as charts, tables, or list reports. Queries built in Explore can be pinned to Storyboards, saved to the Insight Library, or exported.
📍 Product area
Using Explore (entire section)
⚡ Use case
An analyst opens Explore, selects the "Termination Rate" metric, adds "Business Unit" as a dimension, sets the time period to "Last 12 Months," and runs the query to generate a bar chart comparing turnover across divisions.
Related: Query · Metric · Dimension · Insight Library · Pin (Pinning)
-
A map-based data visualization in Storyboards that plots metric values at geographic locations (countries, states, cities, or zip codes). Geomaps display data as sized and colored bubbles or shaded regions on an interactive map, with the granularity determined by the location data available in your data model.
📍 Product area
Using Storyboards → Charts & Visualizations
⚡ Use case
The Facilities team wants to see where remote employees are concentrated. You create a geomap showing headcount by city, with bubble size representing employee count, revealing unexpected clusters in secondary cities.
Related: Storyboard · Tile · Dimension
-
A personal repository where One Model users save queries, charts, and list reports created in Explore for reuse. Saved insights can be inserted into Storyboards, shared via Storyboard permissions, duplicated, renamed, or organized into categories. Each user's Insight Library is personalized.
📍 Product area
Using Explore → Saving & Reusing Insights
⚡ Use case
You've built a complex Explore query showing headcount by cost center with conditional formatting. Rather than rebuilding it each month, you save it to your Insight Library and insert it into multiple Storyboards with one click.
Related: Explore · Pin (Pinning) · Query · Storyboard
-
A table-based report format in Explore that displays detailed, record-level data (as opposed to aggregate metrics). List reports can span multiple data tables, include columns from different entities, support conditional formatting, and be embedded as tiles in Storyboards. They are commonly used for operational reporting, data validation, and employee-level detail.
📍 Product area
Using Explore → Build & Format Reports
⚡ Use case
Your Talent Acquisition team needs a weekly report of all open requisitions with details like job title, hiring manager, days open, and current pipeline stage. You build this as a list report in Explore and pin it to a dedicated Storyboard.
Related: Explore · Drillthrough · Deep Linking · Tile
-
A quantitative measurement defined within One Model that calculates a specific people analytics value. Metrics are built on underlying data tables and include configuration for aggregation type (count, sum, average), time concept (point-in-time, cumulative, average), and optional time functions. One Model provides over 500 out-of-the-box metrics across 14 HR subject areas, and users can create custom metrics through the interface.
📍 Product area
Using Explore → Metrics & Time Concepts
⚡ Use case
"Voluntary Termination Rate" is a calculated metric that divides cumulative voluntary terminations by average headcount. Users can create custom metrics like this in the interface in a few minutes.
Related: Average Headcount · Time Function · Dimension · Cumulative · Point in Time
-
An evolution of the standard Storyboard that organizes content across multiple tabbed pages within a single Storyboard. Pages share a common filter bar (filters applied on one page apply across all pages), and each page can be independently designed, titled, and ordered.
📍 Product area
Using Storyboards → Get Started with Storyboards
⚡ Use case
You build a "Talent Acquisition" multi-page Storyboard with four pages: executive summary, open jobs, candidate pipeline, and DEI in recruiting. Filters for "Business Unit" apply across all four pages.
Related: Storyboard · Tile · Storyboard Library
-
An interactive visualization in Explore and Storyboards that displays organizational hierarchy data as a tree structure. Org charts show reporting relationships, headcount at each level, and key metrics per node. Built from your organizational dimension data and can be saved to the Insight Library.
📍 Product area
Using Storyboards → Advanced Components; Using Explore
⚡ Use case
The CHRO wants to see the SVP-level reporting structure with headcount and span-of-control metrics. You build an org chart in Explore and embed it in an executive Storyboard.
Related: Dimension · Insight Library · Storyboard
-
One Model's comprehensive data platform layer that handles the full data lifecycle: ingestion from source systems, data modeling and transformation, validation, and output to downstream tools. People Data Cloud is the infrastructure that makes raw HR data analytics-ready. It encompasses data sources, the data model, data destinations, and BI connections. (Also marketed as "Data Mesh" in current product naming.)
📍 Product area
People Data Cloud (entire section)
⚡ Use case
When someone says "our data is in People Data Cloud," they mean raw data from Workday, Greenhouse, and other systems has been extracted, cleaned, joined, and modeled into the analytics-ready tables that power Explore and Storyboards.
Related: Data Mesh · Data Source · Data Destination · Entity Relationship Diagram
-
The action of sending a query result from Explore to a Storyboard tile. When you pin a query, the chart or table you've built in Explore becomes a live tile on the selected Storyboard. Pinned tiles maintain their query configuration and update automatically when the underlying data refreshes.
📍 Product area
Using Explore → Saving & Reusing Insights
⚡ Use case
You build a termination trend chart in Explore, click the Pin icon, and select your "Executive Dashboard" Storyboard. The chart appears as a new tile at the bottom of that Storyboard, ready for positioning and styling.
Related: Explore · Tile · Storyboard · Insight Library
-
A dimension placed on the column axis of a query (rather than the row axis), creating a cross-tabulation or matrix view. Pivoting a dimension spreads its values across columns, allowing side-by-side comparison within a single table or chart.
📍 Product area
Using Explore → Explore Basics
⚡ Use case
You query headcount by "Department" (row dimension) and pivot by "Gender" (pivoted dimension). The result is a table where each row is a department and the columns show headcount for each gender side by side.
-
A time concept applied to metrics that counts or sums data as of a specific moment — typically the start or end of a selected time period. Point-in-time measures answer questions like "How many employees did we have on December 31?" They capture a snapshot, not a span.
📍 Product area
Using Explore → Metrics & Time Concepts
⚡ Use case
You need to report headcount to the board as of the last day of each quarter. You configure a "Headcount (End of Period)" metric using the point-in-time concept with an end-of-period filter.
Related: Cumulative · Average Headcount · Time Function
-
An admin feature that allows a designated user to view the One Model instance as if they were another user, seeing exactly what that user sees based on their role-based security permissions. Proxy is used for troubleshooting access issues, validating security configurations, and providing support.
📍 Product area
Admin Toolkit → Identity, Access & User Management
⚡ Use case
An HRBP reports they can't see data for a department. The admin uses Proxy to view the instance as that HRBP, confirms the data is missing, and identifies a missing rule in their data access role.
Related: Role-Based Security · Contextual Security
-
A configured data request in Explore that combines one or more metrics, dimensions, time periods, and filters to retrieve a specific set of results. Queries produce charts, tables, or list reports and can be pinned, saved, or exported.
📍 Product area
Using Explore → Explore Basics
⚡ Use case
"Run a query for Voluntary Termination Rate by Business Unit for the last four quarters" means: open Explore, select the metric, add the dimension, set the time range, and click Run Query.
Related: Explore · Metric · Dimension · Pin (Pinning)
-
One Model's security framework that controls what data, metrics, dimensions, Storyboards, and features each user can access. RBS operates through two types of roles: Application Access Roles (which features a user can use) and Data Access Roles (which data a user can see, including specific metrics, dimensions, columns, and population rules).
📍 Product area
Security & SSO → Authorization & Role-Based Security
⚡ Use case
You create a "Regional HRBP - APAC" data access role that can see all metrics but only for employees in the Asia-Pacific region. You also assign an application access role that allows viewing Storyboards but not editing them.
Related: Contextual Security · Dimension · Proxy · Role Sync
-
An automated process that synchronizes user role assignments between your identity provider (such as Microsoft Entra or Okta) and One Model. When a user's group membership changes in the IdP, their One Model roles update automatically.
📍 Product area
Admin Toolkit → Identity, Access & User Management
⚡ Use case
When a new HRBP is added to the "HRBP-EMEA" group in Okta, Role Sync automatically assigns the corresponding One Model data access role, granting EMEA workforce data access on first login.
Related: Role-Based Security · Contextual Security
-
An admin tool that checks the integrity of Storyboard configurations, identifying issues such as tiles referencing deleted dimensions, metrics that no longer exist, or broken links. Helps admins maintain Storyboard quality after data model changes.
📍 Product area
Admin Toolkit → Admin Tools & Platform Features
⚡ Use case
After a data model update removes an old dimension, you run site validation to find the three Storyboard tiles that still reference it, then fix them before users encounter errors.
Related: Storyboard · Tile · Dimension
-
A small, inline chart (typically a line or bar) embedded within a table cell in a Storyboard tile. Sparklines provide a compact visual trend for each row of data without requiring a separate chart, enabling users to spot patterns directly within tabular reports.
📍 Product area
Using Storyboards → Charts & Visualizations
⚡ Use case
A list report shows departments in rows with their current headcount. A sparkline column next to each department shows the 12-month headcount trend as a tiny line chart, so viewers can instantly see which departments are growing or shrinking.
Related: Change Indicator · Tile · List Report
-
An in-platform SQL query tool for administrators and analysts who need to validate data, run custom queries, or investigate data issues directly against the One Model database. SQL Explorer provides direct access to the underlying data model for advanced troubleshooting and analysis.
📍 Product area
People Data Cloud → SQL & Data Validation
⚡ Use case
During data validation, an analyst uses SQL Explorer to write a query comparing the count of termination records in the raw source table against the processed metric value, confirming they match.
Related: People Data Cloud · Data Mesh · Zero Results / Zero Values
-
One Model's interactive dashboard and data storytelling canvas. Storyboards are composed of tiles (charts, tables, list reports, text, images, geomaps, org charts) arranged on a freely configurable layout. They support filters, drill-through, conditional formatting, and presentation mode. Storyboards are the primary vehicle for sharing insights with business stakeholders.
📍 Product area
Using Storyboards (entire section)
⚡ Use case
A People Analytics team builds a "Workforce Overview" Storyboard containing a headcount trend chart, a turnover rate tile with change indicator, a geomap of employee locations, and a DEI metrics table - all filterable by business unit and time period.
Related: Tile · Multi-Page Storyboard · Storyboard Library · Data Stories
-
The central repository of all Storyboards available to a user, organized by categories. The Storyboard Library is the starting point for finding, opening, creating, copying, and managing Storyboards. Access to individual Storyboards within the library is governed by role-based security.
📍 Product area
Using Storyboards → Get Started with Storyboards
⚡ Use case
A new HRBP opens the Storyboard Library and sees the Storyboards their admin has shared with them: "Workforce Overview," "Talent Acquisition Pipeline," and "DEI Dashboard," organized under category folders.
Related: Storyboard · Multi-Page Storyboard · Role-Based Security
-
An individual visual element on a Storyboard — a chart, table, list report, text block, image, geomap, or org chart. Tiles can be resized, repositioned, styled with custom colors and borders, linked to other Storyboards, and configured with drill-through. Each tile represents one query or content element.
📍 Product area
Using Storyboards → Design & Layout
⚡ Use case
You add a new tile to your Storyboard by pinning a query from Explore. You then style it with a dark header bar, add a title, enable drill-through, and position it in the top-left corner of the page.
Related: Storyboard · Pin (Pinning) · Drillthrough · Annotations
-
A modifier applied to a metric that changes its time context for analytical purposes. Time functions include Rolling Months (looks back N months from the selected date), Time Projected (estimates full-period value based on partial-period data), Year to Date (accumulates from the start of the year), and Annualise (extrapolates partial-year data to a full-year estimate).
📍 Product area
Using Explore → Metrics & Time Concepts
⚡ Use case
You apply a "Rolling 12 Months" time function to your Termination Rate metric so that regardless of which month a user selects, the metric always shows the trailing 12-month rate, smoothing out seasonal variation.
Related: Metric · Alternative Calendar · Point in Time · Cumulative
-
A data modeling configuration that distinguishes between genuinely missing data (true blanks/nulls) and zero values in your dataset. True blanks ensure that analytics correctly exclude missing records rather than treating them as zero, which would distort averages and rates.
📍 Product area
People Data Cloud → Data Modeling & Transformation
⚡ Use case
Some employees don't have a "Salary" value because they're contractors paid hourly. True blanks configuration ensures the "Average Salary" metric excludes these records entirely rather than including them as $0, which would artificially lower the average.
Related: Zero Results / Zero Values · Metric · Allocation
-
A chart type in Storyboards that shows how an initial value is affected by intermediate positive and negative values, leading to a final value. Commonly used in workforce analytics to visualize headcount changes over time — starting headcount, plus hires, minus terminations, plus/minus transfers, equals ending headcount.
📍 Product area
Using Storyboards → Charts & Visualizations
⚡ Use case
An executive Storyboard tile shows starting headcount of 5,000 at the beginning of Q1, +200 hires (green bar), -150 terminations (red bar), +30 transfers in, -20 transfers out, arriving at 5,060 ending headcount.
Related: Storyboard · Tile · Annotations
-
Situations where a query in Explore or a Storyboard tile returns zero values or empty results. Zero results can occur because of data access restrictions (the user's role doesn't permit visibility), because the query produces too many data points to display on screen, or because no data exists for the selected combination of filters. One Model distinguishes between genuine zero values and display limitations.
📍 Product area
People Data Cloud → SQL & Data Validation; Using Explore
⚡ Use case
A user reports seeing zeros in a Storyboard tile for a department they expect to have data. Investigation reveals their data access role doesn't include that department. The zeros are a security restriction, not a data issue.
Related: Role-Based Security · True Blanks · SQL Explorer
One AI & Machine Learning Terms
-
A One AI capability that enriches your existing One Model data with derived or external data. The most common augmentation is Commute Time, which calculates travel times and distances between employee home and office zip codes using an external mapping API (Mapquest). Augmented data becomes available as dimensions and metrics in Explore and Storyboards.
📍 Product area
AI & Automation → One AI Academy → Embedded Insights & Augmentations
⚡ Use case
You run the Commute Time augmentation, and One AI calculates the average commute for each employee based on their home and office postal codes. This data then appears in Explore as a dimension, enabling analysis like "Is commute time correlated with attrition risk?"
Related: One AI · Embedded Insights · Dimension
-
A type of machine learning model in One AI that predicts which category or class a record belongs to. The most common classification use case in people analytics is predicting binary outcomes like attrition risk (will this employee leave: yes/no). Classification models output a probability score for each class.
📍 Product area
AI & Automation → One AI Academy → Modeling with One AI
⚡ Use case
You build a classification model to predict 12-month attrition risk. The model assigns each active employee a probability score (e.g., 0.73 = 73% likely to leave), which is deployed into your data model and visualized on a Storyboard.
Related: Regression Model · Risk Model · Recipe
-
A statistical technique used during One AI model training that splits the data into multiple subsets, training the model on some subsets and testing it on others, rotating through all combinations. Cross validation prevents overfitting by ensuring the model generalizes well to unseen data, not just the data it was trained on.
📍 Product area
AI & Automation → One AI Academy → Machine Learning Foundations
⚡ Use case
When you review your model's Results Summary report, the cross validation scores tell you how consistently the model performs across different slices of your data. A high variance between folds may indicate overfitting to specific subpopulations.
Related: Results Summary Report · Refinement · Upsampling
-
The action of loading a trained One AI machine learning model's predictions into the One Model data model, making the results available as metrics and dimensions in Explore and Storyboards. Deployment makes model outputs operational - visible to users with appropriate permissions.
📍 Product area
AI & Automation → One AI Academy → Interpreting & Evaluating Models
⚡ Use case
After reviewing the Results Summary and confirming the attrition risk model performs well, you click "Deploy" to load the risk scores into the data model. After the next data processing cycle, a "Risk Score" dimension and "High Risk Headcount" metric appear in Explore.
Related: Persist · Results Summary Report · Recipe
-
An automated report generated by One AI before model training that profiles the input data: distributions, missing values, correlations, suspicious variables, and potential data leakage. The EDA report helps you understand your data's quality and suitability for modeling before investing in training.
📍 Product area
AI & Automation → One AI Academy → Interpreting & Evaluating Models
⚡ Use case
Before building a flight risk model, you review the EDA report and discover that "Last Manager Rating" has 40% missing values and a suspicious variable flag. You decide to investigate whether to impute, exclude, or address this gap before proceeding.
Related: Results Summary Report · Feature Engineering · Recipe
-
One AI capabilities that are integrated directly into Storyboard tiles, requiring no separate model setup. Embedded insights include Forecasting (trend projection with confidence intervals), Correlations (statistical significance in scatterplots), and Table Insights (highlighting unusual values in tables). They appear as a lightbulb icon when applicable data is detected.
📍 Product area
AI & Automation → One AI Academy → Embedded Insights & Augmentations
⚡ Use case
A Storyboard tile shows a 24-month headcount trend. The user clicks the lightbulb icon, and One AI overlays a 6-month forecast line with confidence intervals, including seasonality detection with no data science configuration required.
Related: Forecasting · Table Insights · One AI · Storyboard
-
The process of selecting, transforming, and creating input variables (features) that a machine learning model uses to make predictions. In One AI, much of feature engineering is automated, but users can influence which columns are included or excluded and how they're treated through global settings and refinement options.
📍 Product area
AI & Automation → One AI Academy → Machine Learning Foundations
⚡ Use case
One AI automatically creates derived features like "Tenure in Current Role" and "Number of Lateral Moves" from your raw data. During refinement, you notice "Employee ID" was mistakenly included and force-exclude it, since it has no predictive value.
Related: Refinement · Global Settings · SHAP Values
-
An embedded insight capability in One AI that projects future metric values based on historical trends. Forecasting runs in real time within Storyboard tiles, includes confidence intervals, and supports seasonality detection. It does not require a separate model as it operates directly on the time-series data displayed in a chart.
📍 Product area
AI & Automation → One AI Academy → Embedded Insights & Augmentations
⚡ Use case
A Storyboard chart shows 36 months of headcount data. With forecasting enabled, One AI projects the next 6 months as a dashed line with a shaded confidence band, and notes "Seasonality detected: Q4 hiring pattern identified."
Related: Embedded Insights · Storyboard · One AI
-
Configuration parameters that control how One AI processes data and builds models across your instance. Global settings include thresholds for null values, category size limits, dimensionality reduction options, upsampling controls, and default model parameters. They can be overridden at the individual recipe level.
📍 Product area
AI & Automation → One AI Academy → Advanced Configuration
⚡ Use case
Your data has many high-cardinality categorical fields. You adjust the global category size threshold from the default to a higher value so that One AI doesn't automatically collapse categories with fewer members, preserving granularity for your use case.
Related: Recipe · Refinement · Upsampling
-
One Model's integrated artificial intelligence and machine learning platform. One AI encompasses three capabilities: (1) a full automated ML platform for building predictive and causal models (classification and regression), (2) embedded statistical insights within Storyboards (forecasting, correlations, table insights), and (3) data augmentations (commute time). One AI's tagline is "Ethical and Transparent AI in HR."
📍 Product area
AI & Automation (entire section)
⚡ Use case
"We use One AI" could mean: "We built a flight risk model using One AI Recipes," "We enabled forecasting on our headcount Storyboard," or "We ran the commute time augmentation."
Related: Recipe · Embedded Insights · Augmentation · One AI Assistant
-
One Model's conversational AI interface that allows users to interact with their data and the platform using natural language. The Assistant includes multiple modes: Answers (guided Q&A on key workforce questions), Analyze (data exploration via conversation), Chat (general interaction), Visualizations (AI-generated charts), Insights (automated findings), and Storyboard Navigation (natural language navigation to relevant Storyboards).
📍 Product area
AI & Automation → One AI Assistant
⚡ Use case
A CHRO types "What's our voluntary turnover rate in APAC this quarter compared to last year?" into the One AI Assistant. The Assistant retrieves the data, generates a comparison visualization, and provides a natural-language summary of the trend.
Related: One AI · Explore · Storyboard
-
A model deployment option in One AI that "freezes" a trained model so that subsequent data processing runs use the same model (algorithm, features, settings) rather than retraining. Only the input data changes while the model itself remains constant. This ensures prediction consistency over time.
📍 Product area
AI & Automation → One AI Academy → Interpreting & Evaluating Models
⚡ Use case
Your attrition risk model has been validated and approved by stakeholders. You select "Deploy and Persist" so the model doesn't retrain with each data refresh, ensuring consistent risk scores until you deliberately choose to retrain.
Related: Deploy · Recipe · Risk Model
-
One AI's guided model-building workflow that walks users through the process of creating a machine learning model without requiring data science expertise. A recipe defines the question being answered (e.g., "Who is likely to leave?"), the target variable, the population, the time frame, and any configuration overrides. Recipes are the primary entry point for building predictive models in One AI.
📍 Product area
AI & Automation → One AI Academy → Modeling with One AI
⚡ Use case
You create a new recipe: the question is "12-month attrition risk," the target variable is "Is Terminated (Voluntary)," the population is "Active Full-Time Employees," and you set a 12-month prediction window. One AI handles feature selection, model training, and evaluation automatically.
Related: Classification Model · Regression Model · Deploy · Refinement
-
The iterative process of improving a One AI model after reviewing its initial results. Refinement can include forcing features in or out, adjusting global settings, changing upsampling parameters, altering the target population, or modifying the prediction window. Each refinement creates a new model run that can be compared against previous versions.
📍 Product area
AI & Automation → One AI Academy → Advanced Configuration
⚡ Use case
Your initial attrition model includes "Manager Name" as a top feature, which isn't actionable. You force-exclude it and force-include "Engagement Survey Score," then rerun the model. The refined model produces more actionable insights.
Related: Recipe · Feature Engineering · Global Settings · Results Summary Report
-
A type of machine learning model in One AI that predicts a continuous numeric value rather than a category. While classification predicts "yes/no" outcomes, regression predicts magnitudes — such as predicting the number of days until an employee leaves, expected salary, or anticipated performance rating.
📍 Product area
AI & Automation → One AI Academy → Modeling with One AI
⚡ Use case
Instead of predicting whether an employee will leave (classification), you build a regression model that predicts how many months until departure for at-risk employees, enabling more precise intervention timing.
Related: Classification Model · Recipe · One AI
-
A comprehensive report generated after a One AI model training run that provides full transparency into the model's configuration, feature analysis, performance metrics, and any warnings or messages. Paired with the EDA report, the Results Summary gives you everything needed to evaluate whether to trust and deploy a model.
📍 Product area
AI & Automation → One AI Academy → Interpreting & Evaluating Models
⚡ Use case
After training a risk model, you open the Results Summary to review the AUC score, examine which features the model selected, check for data leakage warnings, and understand the estimator type used, before deciding whether to deploy.
Related: EDA Report · Deploy · Cross Validation · SHAP Values
-
A specific application of a One AI classification model that predicts the likelihood of an employee leaving the organization within a defined time window. Risk models are the most common One AI use case and produce per-employee risk scores that can be deployed into Storyboards for operational use.
📍 Product area
AI & Automation → One AI Academy → Modeling with One AI
⚡ Use case
The People Analytics team builds and deploys a 12-month attrition risk model. The resulting risk scores appear on a dedicated "Flight Risk" Storyboard, where HRBPs can filter by department and see which employees are flagged as high risk.
Related: Classification Model · Deploy · SHAP Values · Recipe
-
An interpretability method used in One AI to explain which features (variables) contributed most to an individual prediction and in what direction. SHAP values decompose each prediction into the contribution of each feature, making the model's reasoning transparent and auditable. A critical requirement for ethical AI in HR.
📍 Product area
AI & Automation → One AI Academy → Interpreting & Evaluating Models
⚡ Use case
A manager questions why an employee is flagged as high risk. You review the SHAP values for that individual and see that "3 years in same role" and "below-median pay ratio" were the top contributing factors, providing a clear, explainable rationale.
Related: Results Summary Report · Risk Model · Feature Engineering
-
An embedded One AI feature that automatically applies descriptive statistics to data tables in Storyboard tiles, highlighting potentially noteworthy or unusual values. Table insights add visual cues (such as color-coded cells or indicator icons) that draw attention to outliers, significant changes, or values that deviate from expected patterns.
📍 Product area
AI & Automation → One AI Academy → Embedded Insights & Augmentations
⚡ Use case
A headcount-by-department table in a Storyboard automatically highlights three departments in amber because their turnover rates are statistically higher than the organizational average, drawing the viewer's eye without requiring manual threshold configuration.
Related: Embedded Insights · Storyboard · One AI
-
A technique used in One AI model training to address class imbalance — when one outcome (e.g., "stayed") vastly outnumbers another (e.g., "left"). Upsampling creates synthetic copies of the minority class so the model doesn't simply learn to predict the majority class for everything. It can be toggled in One AI's global or recipe-level settings.
📍 Product area
AI & Automation → One AI Academy → Advanced Configuration
⚡ Use case
Your organization has a 5% annual voluntary turnover rate, meaning 95% of training records are "stayed" and only 5% are "left." You enable upsampling so the model adequately learns attrition patterns rather than defaulting to "stayed" for everyone.
Related: Global Settings · Recipe · Refinement · Cross Validation
This glossary is maintained as part of the One Model Help Center. Last updated: February 2026. For corrections or additions, contact the One Model content team.
Comments
0 comments
Please sign in to leave a comment.