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.

Try a question

    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 agentAnsweredPeople

    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

    • Your access All locations

    What happened meets
    what could happen

    Customer Support lost 75 agents in the last 12 months, more each quarter, most in their first year. TrenData carries that forward: the amber fan is every path the next 12 months can take, one for each mix of raise, mentoring and hiring plan.

    Agents who left Customer Support, added up since October 2025: 75 by today, with exits rising each quarter. From today the total fans out depending on the decision: 30 paths for different raises, mentoring and hiring plans reach between 129 and 156 by next September; doing nothing reaches about 153. The shaded range widens the further it looks. Sample data from a simplified workforce model.

    Agents who left, added up30 decisions, each a pathDoing nothingSample data from a simplified workforce model

    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.

    It starts as one row in a file.

    Upload CSV, Excel or JSON files, or connect a REST API on a schedule. Here a resignation in Customer Support, Austin, arrives with 48,209 other employee movements.

    A CSV file of 48,210 employee movements. One row is highlighted: employee E-20417, Customer Support, Austin, resignation, effective 14 August 2026.

    Checked before it counts.

    TrenData profiles every column, maps it to a shared model and checks the dates. Nothing reaches a dashboard until an approver signs off.

    The same row as a column of checked fields: ID unique, department and location mapped, event one of four known values, dates consistent. Approved by Maria R., HR operations.

    Counted by a definition everyone shares.

    The resignation is counted by the certified voluntary turnover metric: a written formula with a version history, used by dashboards and AI answers alike.

    Voluntary turnover rate, version 3: 48 voluntary separations, including this one, divided by 1,262 average headcount, times 100, is 3.8% for the third quarter.

    Explained in plain language.

    The chart shows the third quarter rising, and TrenData says why. Ask a follow-up and keep going until you reach the cause.

    Voluntary turnover by quarter: 3.1%, 2.8%, 2.9%, 2.9%, then 3.8% in Q3 2026. Most of the increase came from Customer Support in Austin and Denver, where resignations in the first year doubled.

    Carried forward, with its uncertainty.

    Workforce models forecast headcount and show where retention risk is building. The range widens the further it looks, and each model shows how recent it is.

    Customer Support headcount: 412 today. If exits continue at the current pace, about 398 by March, with a likely range of 386 to 410.

    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.

    People and HR

    Headcount, turnover, pay, absence and learning, with forecasts for the workforce you will need.

    Voluntary turnover, last 8 quarters3.8%

    Pick a question below to see its answer here.

    Also asked in this pack: Do people in the same role earn comparable pay?

    Recruiting

    Requisitions, applications and hires, from the first posting to the accepted offer.

    Candidates by stage, last quarter52 hires

    Pick a question below to see its answer here.

    Also asked in this pack: Which sources bring the most hires?

    Supply chain

    Purchase orders, inventory, shipments and supplier quality, measured with 80 ready-made metrics.

    On-time in-full, last 6 months93.0%

    Pick a question below to see its answer here.

    Also asked in this pack: Where is freight cost per order rising?

    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. Query the AI wrote
      select sum(voluntary_separations)
           / avg(headcount) * 100
      from   workforce_q3_2026
      where  region = 'North America'

      The last line is added by the access rule. The AI cannot leave it out.

    Answer

    3.9% 16 of 415 people left

    LocationRegionHeadcountLeft
    Austin
    North America
    180
    8
    Denver
    North America
    140
    6
    Toronto
    North America
    95
    2

    Ana can read 3 of 6 locations. The rest never reach the query.

    Audit log

    1. Ana asked, voluntary turnover rate version 3, region North America, answer 3.9%

    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