- Industry
- Insure AI
- Department
- Engineering
- Team
- Application Engineering
- Kind
- Decision-maker
- Approver
- Tech Lead
- Code
- ENG-04
It reviews code changes to the insurance platforms for bugs and standards and decides approve, request changes or block. It looks hardest at changes to rating, claims payment, reserving feeds and health data handling, where a bug means mispriced policies, wrong payouts or a data breach.
How it works
7 steps, and a person signs off before step 7.
- 01Looks it up
Finds the record the request is about in the code repository: the change diff, its tests and the history of the files touched..
- 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.
- 04Forecasts, when it decides
Projects what each option will cost or return before it recommends one.
- 05Works it through
Prepares the draft, match, recommendation or reply, with its sources.
- 06Asks a person
Hands the work to Tech Lead with its reasoning before anything goes out.
- 07Acts
Once approved, it acts in the code repository: it posts the review with its decision and comments. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The coding standards, the sensitive paths (rating, claim payments, reserving and IFRS 17 feeds, health data, sanctions checks), the review checklist and known failure patterns.
Reads from
the code repository: the change diff, its tests and the history of the files touched.
Acts in
the code repository: it posts the review with its decision and comments.
Hands back
An object with decision (approve, request_changes or block), findings (a list, each with file, line, severity, issue and fix) and summary.
Asks a person when
The change touches rating/, claims/payments/, reserving/, ifrs17/, health/, sanctions/ or a finding is at high or above.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Block on findings of | high | At or above this, block. |
| Most lines in one review | 600 | Larger changes are split. |
| Require tests | On | When on, logic changes need tests. |
| Sensitive paths | rating/, claims/payments/, reserving/, ifrs17/, health/, sanctions/ | Comma-separated. |
| Most comments | 15 | Keep the review focused. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- Block on findings at high or above.
- Changes in rating/, claims/payments/, reserving/, ifrs17/, health/, sanctions/ need a second reviewer, the Tech Lead.
- When on is true, request tests for logic changes.
- Ask for a split above 600 lines; keep to 15 comments.
- Flag any rounding, currency or date logic in premium or claim amounts, any health data logged, and any skipped sanctions check.
See the Code Review Agent agent on your own work.
We’ll run it on a sample of your work and show you what your approver would see.