Conversational Assessments
A conversation with every employee,
not a static survey.
A Conversational Assessment is an AI-led interview, built from your organisation's own workforce fingerprint, that adapts its questions to what each person actually says. It combines scaled items with genuine two-way dialogue, so you keep the trend data a survey gives you while also learning why the score is what it is.
Why this exists
Why isn't a survey enough?
Most strategies that fail to land share the same root cause. Leadership didn't actually know what was happening at the front line. Not because nobody was asking — because the tools available only ever asked the same fixed question of everyone, and a fixed question cannot follow up.
When someone rates communication poorly on a five-point scale, a survey stops there. It never asks what "unclear" actually meant, or what would have made it clear. That follow-up is where the useful information sits, and it's exactly what a static instrument cannot capture.
How it works
How does a conversational assessment run?
It starts with a scale, and follows up like a person would.
- 1
Built from your context
Every assessment is generated from your organisation's own workforce fingerprint — its structures, roles, hazards and history — rather than a standard template. Two organisations receive materially different questions, because they are starting from materially different pictures.
- 2
Scaled items, to start
Each conversation opens with Likert-style scaled questions, chosen for the area being explored. These preserve the benchmark and trend data an organisation already relies on, and they shape what the conversation asks next.
- 3
A genuine follow-up, not a script
Where a response indicates something worth exploring — a low score, an unexpected answer — the agent asks a further question generated from that specific response, not a fixed decision tree. If someone says communication has been unclear, the next question explores what “unclear” meant to them.
- 4
Every conversation is its own
Because follow-ups are generated individually, no two people have the same conversation. Each is shaped by what that person has said, in their own words, at their own pace.
Where it goes
What happens to what people say?
Every conversation feeds directly into your organisation's workforce fingerprint — the live model Gentia reasons across to show where risk is building and why. Personal details are removed before any analysis, and reporting is produced at team, group or organisational level.
That means consultation isn't a separate exercise sitting beside your risk picture. It becomes part of it. A theme raised in a conversation this week can sharpen a finding your organisation already holds, or surface one that hadn't shown up anywhere else.
What you get
What does the organisation receive?
Anonymised, aggregated reporting
Personal details are removed before analysis. Reporting is produced at team, group or organisational level — never at the individual level.
Themes, not just scores
Because the conversation goes beyond a number, the reporting explains what is driving a score, not only what the score is.
Direct input to the fingerprint
Findings connect straight into your organisation's workforce fingerprint, sharpening the picture rather than sitting apart from it.
A route to action
Themes that indicate a specific, addressable condition are flagged for control design and action — not filed as a data point to revisit next year.
Multi-rater
Can this run as a 360?
Yes. The same conversational approach can be built as a multi-rater or 360 assessment.
Peers, direct reports and managers can each be asked questions relevant to their vantage point, generated from the same organisational context, with the same adaptive follow-up. The result is a 360 that captures what people actually observed, in their own words, rather than only where they landed on a scale.
How it fits
Does this replace our existing survey?
No. It's designed to sit alongside it.
Conversational Assessments are built to complement the instruments an organisation already uses, not replace them. Scaled items are retained inside the experience specifically so existing benchmark and trend data continues, while the conversational layer adds the context that explains it. If your organisation runs an annual engagement survey, this is not an argument to stop.
Personal details removed before analysis
Names and identifying details are stripped before any AI processing.
Reported at team level, never individual
No output identifies a person, and small groups are protected from being singled out.
Not a performance or disciplinary tool
Assessments explore work design and conditions, not individual performance.
Built on 14+ years of practice
Every question set draws on Gentia's expert-labelled dataset, mapped to ISO 45003 and Australian work health and safety regulation.
Where can people take it?
Conversational Assessments are currently delivered through the browser. Integration with Microsoft Teams and Slack is planned.
FAQ
Conversational Assessment questions, answered.
A Conversational Assessment is an AI-led interview built from an organisation's own workforce fingerprint. It combines scaled questions with adaptive follow-up questions generated from each person's own responses, so it captures both trend data and the context that explains it.
A survey asks a fixed question of everyone and stops at the answer. A Conversational Assessment follows up on what a person actually says — if someone rates communication poorly, the next question explores what that meant to them, rather than moving to the next fixed item on the form.
No. Because follow-up questions are generated from each person's own responses, no two conversations are identical. Everyone starts from the same scaled items, relevant to the area being explored, but the conversation that follows is individual.
No. It is designed to complement your existing instruments rather than replace them. Scaled items are retained inside the conversation specifically to preserve the trend and benchmark data your organisation already relies on.
Yes. The same conversational approach can be structured as a multi-rater or 360 assessment, with peers, direct reports and managers each asked questions relevant to their vantage point.
Personal details are removed before any analysis, and reporting is produced at team, group or organisational level. No output identifies an individual, and small groups are protected from being singled out.
No. Conversational Assessments explore work design and conditions, not individual performance, and Gentia's Acceptable Use Policy prohibits using the platform for performance management or disciplinary purposes.
Every conversation feeds directly into the organisation's workforce fingerprint — the live model Gentia reasons across — as well as into anonymised, aggregated reporting.
The agent can adapt to the language an employee uses to respond, so consultation is not limited to a single-language rollout.
It isn't. Gentia does not train its models on client data.
Currently through the browser. Integration with Microsoft Teams and Slack is planned.
Hear what your people are actually experiencing.
Start with a Discovery Brief. A short conversation, plus whatever you already have, and you get back a brief on what matters in your organisation and where consultation would tell you the most.
Nothing to prepare. No survey to run.