TrenData
Menu

Ask a business question
Get an answer you can check

TrenData brings your people, recruiting and operations data together, measures it with metrics your team agrees on, and answers in plain language with the working shown.

Ready today for HR, recruiting and supply chain data. For other industries, we build the first set of metrics with you.

Ask it something

Pick a question and watch how TrenData gets to the answer: which metric it chose, whose access applied and how many records it read.

People

Recruiting

Supply chain

Sample data. In TrenData, answers come from your approved datasets and certified metrics.

Ask your agentPeople

Why did voluntary turnover rise in Q3?

  1. Matched “voluntary turnover” to a certified metric
  2. Applied your access: all locations
  3. Read 48,210 employee movement records
  4. Compared Q3 with the four quarters before it

Voluntary turnover rose from 2.9% to 3.8%. Most of the increase came from Customer Support in Austin and Denver, where resignations in the first year doubled.

Voluntary turnover rate by quarter

Metric
Voluntary turnover rate, version 3
Data
Employee movements, approved 2 October
Your access
All locations

Make a decision
Watch the next 12 months move

Customer Support is losing first-year agents. Try a raise, mentoring or a different hiring plan and see what each choice costs and saves.

Your decision

2%
Hiring plan
Agents leaving in 12 months
6878 if you do nothing
Net effect after costs
+$20Ksavings minus what it costs
Staffing
On target430 agents from January
Agents who leave, added up month by month
Cumulative exits over the next 12 months.
Doing nothing 78Your decision 68Likely range
Agents on the floor against the 430 you need from January
Agents on the floor each month.

With a 2% raise for first-year agents, about 10 fewer agents leave over the next 12 months. Replacement savings cover the cost, leaving $20K ahead. You reach the 430 agents you need in January and stay there.

See the numbers and assumptions
Month by month forecast for your decision
MonthAgents on the floorExits so farExits so far, doing nothing
Oct41267
Nov4131214
Dec4121820
Jan4302327
Feb4302934
Mar4303540
Apr4304147
May4304653
Jun4305260
Jul4305766
Aug4306372
Sep4306878

Assumptions for this sample: 412 agents today, 35% in their first year. First-year agents leave at 3.2% a month, others at 0.9%. Replacing an agent costs $12,000, a first-year salary is $42,000, and mentoring costs $40,000 a year plus $500 per new hire. New hires take two months to arrive. Raises have diminishing returns.

A simplified model with sample data, made to show how decisions and forecasts connect. In TrenData, forecasts come from your own records and show how recent the model is.

Behind every answer
five steps you can inspect

TrenData does the preparation before it answers, so the answer holds up when someone asks where the number came from.

Connect dataEmployee movements
employee_movements.csv48,210 rows
  1. ProfiledColumns, types and empty values
    Done
  2. Fields mapped23 of 23 fields matched
    Done
  3. Dates checkedEffective dates are consistent
    Done
  4. ApprovalMaria R., HR operations
    Waiting
After approval, an API source can refresh on a schedule.

Clean data before clever answers.

Upload CSV, Excel or JSON files, or connect a REST API on a schedule. TrenData profiles each dataset, maps it to a shared model and waits for an approver before anything reaches a dashboard.

  • Files and APIs
  • Quality checks
  • Approval before go-live
Metrics libraryHR analytics
  • Headcount
  • Voluntary turnover rate
  • Average base pay
  • Open requisitions
  • Absence days

Voluntary turnover rate

Voluntary separations ÷ average headcount × 100
Version
3, compared with version 2
Counts
Resignations only
Used by
Dashboards and chat answers

One definition for every number.

Each metric has a written formula and a version history. Dashboards and AI answers use the same definition, so two teams don’t bring two different numbers to the same meeting.

  • Metrics library
  • Versioned formulas
Workforce overviewDashboard
Location: allDepartment: allLast 12 months
Headcount1,284
Voluntary turnover3.8%
Open requisitions46

Headcount by department

Drill path: all departments, Customer Support, Austin

A dashboard on day one.

For supported datasets TrenData builds a starting dashboard for you. Filter it, drill from a department down to the records behind it, and add any metric from the library.

  • Generated dashboards
  • Filters
  • Drill-down
Average days to fill, engineeringRecruiting
What changed
  • Engineering roles took 52 days to fill in September, up from 38 in June.
  • Fewer candidates moved past the first interview.
  • Roles filled through referrals closed 11 days faster.

Every chart can explain itself.

Ask why a line moved and get a short explanation in plain language. Keep going with the assistant beside the chart until you reach the cause.

  • Chart insights
  • Drill assistant
Headcount forecastModel updated 3 days ago
ActualForecastLikely range
Retention risk by groupLevel
Customer Support, first yearHigh
Engineering, two to four yearsMedium
Sales, five years or moreLow

Look ahead, not only back.

Workforce models forecast headcount and time to hire, show where retention risk is building and check pay equity. Each model shows how recent it is before you rely on it.

  • Forecasts
  • Retention risk
  • Pay equity

Illustrative screens with sample data, not customer results.

Built around the questions
your teams already ask

Each industry pack arrives with its data model, metrics and dashboards, so you start from a working model instead of a blank page.

Working in another industry?

Construction, fitness, legal or energy: the same foundation adapts. We shape the data model and metrics with your team, then you ask the questions.

Talk about your industry

Answers people can stand behind

An AI answer is only useful if a manager can repeat it in a meeting. Watch how one answer is made, then switch who is asking.

Who is asking?
  1. Question

    What was voluntary turnover last quarter?

  2. Certified metric

    Voluntary separations ÷ average headcount, version 3

  3. Access rule

    Region is North America

  4. Answer

    3.9% 16 of 415 people left

LocationRegionHeadcountLeft last quarterAccess
AustinNorth America1808IncludedNot in access
DenverNorth America1406IncludedNot in access
TorontoNorth America952IncludedNot in access
MadridEurope1203IncludedNot in access
LisbonEurope852IncludedNot in access
BerlinEurope1103IncludedNot in access

Sample data. The same question, asked by two people with different access, reads different rows and gets a different answer. Both are logged with the metric version they used.

Show the working
Every answer names the metric, the data and the filters it used, and keeps a trace of each step.
Access rules go into every query
People see only the rows they are allowed to see, whether they open a dashboard or ask the assistant.
Roles fit the job
Admins connect data and manage the workspace. Everyone else explores, asks and shares what they find.
A record of every question
Each question is logged with the dataset, metric and version that answered it.
Details for IT and security reviewers
  • Sign in with email and password or with Google.
  • Four roles: global admin, admin, user and viewer.
  • Row-level access policies on dataset fields, added as mandatory filters to generated queries before they run.
  • Analysis code produced by the assistant is checked by a validator before it runs.
  • Each dataset load is profiled, approved, promoted and audited.
  • Answers are generated with OpenAI models through TrenData’s AI service.

This page makes no security certification claims. Ask us for current security documentation.

Bring your questions
We’ll bring the data model

In a demo we walk through TrenData with sample data from your industry and the questions your team asks most.

Book a demo