Applied products
FlightRisk
Employee attrition intelligence
Attrition risk with the explanation attached. Open it with a one-click demo persona — CHRO, VP or manager — and see the whole thing without an account.
What FlightRisk does
FlightRisk scores how likely each employee is to leave and shows why. The executive view opens on total headcount, the number of employees at risk, the cost exposure that risk represents and the average risk score, then states what the data shows in plain sentences: which department has the highest attrition rate, and how much more likely employees who work overtime are to leave than those who don’t.
A predictive-alerts list flags the employees most likely to leave in the next 30 days. Each alert names the pattern behind it and a prescribed action. For a burnout profile — overtime, poor work-life balance and a high risk score together — the action is to cap overtime that week, reassign at least one project and book a well-being one-to-one.
Around the dashboard are the working views:
- Department Explorer and Manager View
- Timeline, Trends and Benchmarking
- What-if Analysis and a Cost Calculator
- An Action Tracker for the interventions
- Upload Data
Who it’s for
HR leaders and the managers who act on retention. The demo opens as one of three personas, with no account: a CHRO who sees the whole organization, a VP of Sales scoped to one department, and a manager who sees only their direct reports. Switching between them shows how role-based access narrows what each person can see.
The data in the demo
The demo runs on a sample dataset of 1,470 employees and is labelled as sample data in the app; the employee names in it are synthetic. Demo mode stores data locally in your browser.
How it’s built
FlightRisk is the product version of Analytica’s employee-turnover method, which runs in four stages: measure turnover and what it moves with, find the features that drive it, build a model that predicts who will leave, and recommend how to keep them.
The method segments the workforce into groups with similar characteristics before modelling, because one model across a diverse population gives less specific results. It also engineers signals an HR system doesn’t store, such as a job-hop index, compa-ratio, and the age gap between an employee and their manager.
What it isn’t
The demo is not a client’s workforce. Every figure in it, from the headcount to the cost exposure, comes from the sample dataset.
