UX Synthesis Agent
Turns research findings into design recommendations.
- Industry
- Insure AI
- Department
- Design
- Team
- Product Experience
- Kind
- Investigator
- Approver
- Head of Distribution
- Code
- DSN-02
It turns usability research on insurance journeys, such as abandoned quotes, confusing FNOL steps or broker portal workarounds, into findings and design recommendations, weighing competing explanations from interviews, session recordings and journey data. It separates design causes from pricing and underwriting causes. **The problem it solves:** Why customers struggle with the journey studied: where they drop off or misunderstand, and whether design, wording, price or underwriting questions explain it.
How it works
6 steps, and a person signs off before step 6.
- 01Looks it up
Finds the record the request is about in the policy administration system: quote starts, step drop-offs and binds for the journey and period studied..
- 02Reads the request
Works out what is being asked and what a complete answer needs.
- 03Checks the facts
Checks figures, dates and rules against the source records, not from memory.
- 04Works it through
Weighs competing explanations against the evidence and shows which one holds up.
- 05Asks a person
Hands the work to Head of Distribution with its reasoning before anything goes out.
- 06Acts
Once approved, it acts in the agent and broker portal: it publishes the synthesis to the research repository and links it to the journey. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The research repository, earlier insurance journey studies, the severity scale, and journey analytics definitions (quote drop-off, FNOL completion, self-service rate).
Reads from
the policy administration system: quote starts, step drop-offs and binds for the journey and period studied.
Acts in
the agent and broker portal: it publishes the synthesis to the research repository and links it to the journey.
Hands back
An object with findings (a list, each with finding, severity, evidence, participants and competing_explanations), recommendations (a list, each with recommendation, finding and effort) and open_questions.
Asks a person when
A finding shows customers misunderstanding cover, exclusions or their cancellation rights.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Participants for a pattern | 3 | An observation from fewer participants is an anecdote. |
| Severity scale | 1-4 | The scale for findings. |
| Most recommendations | 8 | The report keeps the highest-severity ones. |
| Anonymise quotes | On | When on, names, policy numbers and claim or health details are removed from quotes. |
| Findings that need two sources | pricing confusion, trust in claims, disclosure misunderstanding | Comma-separated. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- A pattern needs at least 3 participants.
- Rate severity on the 1-4 scale; a finding where customers misunderstand cover or exclusions is always top severity.
- Findings on pricing confusion, trust in claims, disclosure misunderstanding need research and journey data agreeing.
- Separate design causes from price and underwriting causes; a drop-off at the price step is not a design finding without evidence.
- Keep at most 8 recommendations; when on is true, remove names and claim or health details.
See the UX Synthesis Agent agent on your own work.
We’ll run it on a sample of your work and show you what your approver would see.