Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for insurance data analysts

Mortality Data Quality Cleanup Agent

Produce a clean study dataset with documented fixes and exclusions.

Data cleanupQuality controlFinanceInsurance
See how it worksGet access
Mortality Data Quality Cleanup Agent: what goes in, what the agent does and what you get

What it does

A mortality study built on policy and claim data with impossible dates, duplicate records and mismatched statuses gives wrong death rates. This agent scans the policy and claim files for errors such as death before issue, birth dates in the future, duplicate policies and claims without a matching policy. It proposes a fix for each error type where the right answer is clear, and leaves the rest as exclusions. After applying proposed fixes in a working copy, it reruns the checks and measures the error rate. It repeats until the rate falls below the limit or no further fix is clear, and reports exclusions with counts. The actuary approves the data used.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueApprovedNo 1 STARTS WHEN Experience study begins 2 USES A TOOL Load policy and claim files 3 DOES Run checks for impossible dates, duplicates andorphan claims 4 DOES Propose a fix for each error type with a clearanswer 5 DOES Apply fixes to a working copy 6 USES A TOOL Rerun the checks and compute the error rate 7 CHECKS THE RESULT Is the error rate below the allowed limit? If not: look for further fixes from other records, thenexclude what cannot be fixed. Back to step 4. 8 DOES Compare exposure and deaths before and aftercleaning 9 DOES Write the exclusion report with counts by reason 10 YOU APPROVE Actuary approves the data used 11 RESULT Clean dataset and exclusion log
Read the steps as a list
  1. Experience study begins
  2. Load policy and claim files
  3. Run checks for impossible dates, duplicates and orphan claims
  4. Propose a fix for each error type with a clear answer
  5. Apply fixes to a working copy
  6. Rerun the checks and compute the error rate
  7. Is the error rate below the allowed limit?If not: look for further fixes from other records, then exclude what cannot be fixed. Back to step 4.
  8. Compare exposure and deaths before and after cleaning
  9. Write the exclusion report with counts by reason
  10. Actuary approves the data usedThe agent waits here for your OK.
  11. Clean dataset and exclusion log

How it decides

It fixes an error only when the correct value can be taken from another record, and otherwise excludes the record and counts it.

  • Accept an error rate below 0.5 percent of records
  • Fix a date only if another system holds the same date
  • Exclude a record when two sources disagree and no tie-break exists
  • Flag any fix that changes more than 1 percent of exposure

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Error rate limit (default 0.5 percent)
  • Checks to run
  • Fields used as tie-breakers
  • Exclusion report format

What keeps you in control

It always asks you first

  • Actuary approves the final dataset
  • Data owner approves fixes to source records

Hard limits

  • Works on a copy, never the source data
  • Lists every fix and exclusion

It stops when

  • Done: error rate under the limit and report approved
  • Stop: more than 5 percent of records excluded, review the source

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe agent scanned 1.84 million policies. It found 2,310 with death before issue date, 4,880 duplicates and 960 claims without policies. Matching on policy number fixed 3,900 duplicates. Error rate was still 0.9 percent, above the 0.5 limit, so it checked a second system and fixed 1,100 more dates. Rate fell to 0.4 percent. The actuary approved.

More agents for insurance data analysts

Data cleanup · Analysis

Loss Run Analysis Agent

A clean, checked loss history with frequency, severity and large losses

10 steps2 checks3 jobs
Quality control · Analysis

Survey Weighting and Sample Check Agent

Produce weighted survey estimates that are stable and well documented.

10 steps2 checks5 jobs
Analysis · Risk management

Mule Account and Network Link Analysis Agent

Find the accounts linked to a confirmed fraud account and rank them by strength of evidence.

10 steps1 check3 jobs
Decisions · Analysis

Mortality and Lapse Assumption Update Agent

Propose supported assumption updates with reserve impact, ready for review.

11 steps2 checks2 jobs