SQL · Find the facts
The question
Which policy covered this vehicle on the loss date?
Select the claim, loss date and policy version.
Explore SQLInsurance
Find the evidence
for each claim.
Retrieve the policy in force, relevant clauses and supporting documents for a claim. Help handlers identify what is missing before reviewing coverage.
Motor claim review
Find repair evidence and the policy terms for a motor claim.
Illustrative use case01 / Use case guide
10 practical use cases.
Compare the data models behind everyday industry tasks, from finding records to following connections.
SQL Filter and calculate Vector Match similar meaning Graph Follow relationships Full-text Find words and phrases
Suggested model combination
Scroll across to compare all four models| Use case | How the data models help | SQL | Vector | Graph | Full-text | Source |
|---|---|---|---|---|---|---|
| Claims coverage evidence review Find the policy wording and endorsements relevant to a reported loss. | Use SQL to select the policy version on the loss date and Full-text to locate exact clauses for the handler. | SQL included | Vector not needed for this example | Graph not needed for this example | Full-text included | |
| Claims intake document routing Associate incoming evidence with the correct claim and policy. | Use SQL to filter open claims, Graph to follow recorded claim–policy links, and Full-text to locate supplied claim references. | SQL included | Vector not needed for this example | Graph included | Full-text included | |
| Medical claim document review Find relevant passages across a claim's medical documents. | Use SQL to scope the claim and document dates, Vector to retrieve related descriptions, and Full-text to find exact medical terms. | SQL included | Vector included | Graph not needed for this example | Full-text included | |
| Underwriting submission evidence review Bring exposure data, loss histories and risk reports into an underwriter's review. | Use SQL to select submission records, Vector to retrieve related risk narratives, and Full-text to locate exact loss references. | SQL included | Vector included | Graph not needed for this example | Full-text included | |
| Connected claims investigation support Examine recurring parties across claims referred for investigation. | Use SQL to scope referred claims by date and Graph to trace documented claimant, vehicle and provider connections for investigators. | SQL included | Vector not needed for this example | Graph included | Full-text not needed for this example | |
| Delegated risk data checks Investigate missing or inconsistent fields in coverholder risk reports. | Use SQL to surface missing fields and duplicate references, and Full-text to find the corresponding binder wording for review. | SQL included | Vector not needed for this example | Graph not needed for this example | Full-text included | |
| Premium bordereau reconciliation Trace premium reporting discrepancies to the relevant binder and reporting party. | Use SQL to compare reported premium records by period and Graph to trace their binder, coverholder and managing-agent relationships. | SQL included | Vector not needed for this example | Graph included | Full-text not needed for this example | |
| Delegated claims authority review Verify the administrator's recorded appointment before routing a claim. | Use SQL to filter appointment records, Graph to follow administrator–managing-agent relationships, and Full-text to retrieve the documented scope of authority. | SQL included | Vector not needed for this example | Graph included | Full-text included | |
| Catastrophe forecast evidence checks Explain differences between catastrophe forecast returns and supporting scenarios. | Use SQL to compare reported snapshots, Graph to connect returns to supplied model-run records, and Full-text to retrieve scenario assumptions. | SQL included | Vector not needed for this example | Graph included | Full-text included | |
| Catastrophe loss reporting support Assemble claims associated with a named catastrophe for reporting review. | Use SQL to select reporting periods, Graph to follow recorded claim–event links, and Full-text to find the catastrophe's official description. | SQL included | Vector not needed for this example | Graph included | Full-text included |
Use the query interfaces you need, then connect their results through shared record IDs in your application.
Industry sources 8
These sources describe real industry workflows. The model combinations above are proposed OriginChainDB designs, not customer deployment claims.
- Allianz — Smarter claims management, smoother settlements (opens in a new tab)
- Zurich Insurance — AI at Zurich: medical document intelligence and claims intake (opens in a new tab)
- Zurich North America — AI tool brings real-world value to Zurich USMM underwriters (opens in a new tab)
- National Insurance Crime Bureau — Staged and Caused Accident Investigation Guide (opens in a new tab)
- LIMOSS — Delegated Data Manager: binder and bordereaux management (opens in a new tab)
- Lloyd's — Delegated Claims Administrators: appointment and oversight (opens in a new tab)
- Lloyd's — 2027 Capital and Planning: Catastrophe Risk (opens in a new tab)
- Lloyd's — Catastrophe Codes and Major Loss Reporting (opens in a new tab)
02 / A closer look
Checking a repair estimate against the policy
A repair estimate arrives after a claim has been opened. The handler must check the policy version in force, find the relevant clause and identify which evidence is still missing before making a decision.
What makes it difficult
- 01
Policy wording and endorsements live in separate files.
- 02
The same damage is described differently in each document.
- 03
An incomplete evidence pack causes another follow-up.
Suggested approach
Match the claim to the policy version
Use the loss date to select the applicable policy. Search its clauses and linked claim documents, then show missing evidence to the handler.
03 / Queries
What each data model does
Select a model to see the question it answers.
Vector · Match the meaning
The question
Which passages describe damage like this repair report?
Retrieve similar descriptions from approved claim documents.
Explore Vector Graph · Follow the connections
The question
Which documents and parties belong to this claim?
Follow declared links between claim, policy, vehicle and repairer.
Explore Graph Full-text · Find the exact wording
The question
Where does the policy mention the excess or exclusion?
Locate exact clause names, reference numbers and terms.
Explore Full-textUse shared record IDs to combine results from the different query interfaces in your application.
04 / Implementation
How to build it
- STEP 01
Choose one claim type
Use authorized closed claims with unnecessary personal data removed. Have handlers identify the documents and clauses required for each review.
- STEP 02
Link claims and documents
Assign claim, policy and passage IDs. Store effective dates and file references. Your ingestion pipeline extracts text and maintains search indexes and embeddings.
- STEP 03
Retrieve within the claim
Filter by access and policy dates. Search relevant wording and follow document links. Show the source passages and missing inputs in the handler’s screen.
- STEP 04
Compare with handler reviews
Check policy versions, missing evidence and incorrect matches against closed cases. Pilot with handlers while keeping claim decisions in the existing approval process.
05 / Testing
What to measure
Compare these results with your current process before rollout.
Evidence completeness
Compare the retrieved pack with the documents handlers marked as required.
Correct policy version
Check whether each cited clause belongs to the policy effective on the loss date.
Review effort
Compare time spent finding evidence against the existing workflow, without assuming a saving.
06 / Sources
Examples from the industry
Published projects and reference material relevant to this use case.
Allianz
Insurance Copilot supports claims handlers
Allianz describes an Insurance Copilot launched for automotive claims in Austria, helping adjusters ask questions and navigate case information.
Read the sourceZurich Insurance
Document intelligence and claims intake
Zurich identifies medical document intelligence and automated claims intake among its insurance AI use cases, alongside responsible AI practices.
Read the sourceThese are industry references, not OriginChainDB customer stories. The implementation above is a proposed example.
Build with OriginChainDB