A life insurance policy sold in 1998 might still be active today. The customer may claim it in 2040. Somewhere in that insurer’s system sits the original proposal form, the medical report, the underwriting notes, every premium receipt, every endorsement, and the nomination record that was updated three times along the way. Now multiply that by a few million policies.
That’s the shape of the problem in Indian insurance. Industry old data becomes genuinely useless after a few years, insurance data stays legally and commercially relevant for decades. A general insurer might close a motor claim in eighteen months. A life insurer is holding records that will outlive the people who wrote them. And the volume keeps climbing.
Why Retrieval Is the Real Pain Point
The IT head says storage capacity is not the issue because disks are affordable.The issue is what happens when someone actually needs something out of it. A claims officer needs the 2009 policy document. An IRDAI audit asks for a sample of underwriting files from a specific period. A legal team is defending a repudiated claim and needs the original medical questionnaire. In many insurers, that request kicks off a manual hunt. Someone checks the current system. Not there. Check the archive. Partially there. Calls the vendor who managed the legacy platform before the 2016 migration. Waits.
What IRDAI Expects
IRDAI guidelines require insurers to maintain policy records, claims documentation and related data for defined periods, with policy records generally retained through the life of the contract and for a specified period after settlement or termination.
But retention alone doesn’t satisfy a regulator. The data has to be produceable. Auditable. Traceable. An insurer that has technically kept the records but cannot retrieve them within a reasonable timeframe is not in a strong position during an inspection.
The exact retention periods vary by product line and record type, and they’ve been revised over the years, so these should be verified against current IRDAI circulars rather than assumed from internal practice.Â
Building an Archive That Actually Work
Tier the Data by How It’s Used, Not How Old It Is – Age is a poor proxy for access frequency in insurance. A twenty-year-old life policy that’s still in force gets pulled up more often than a three-year-old motor policy that expired last renewal cycle.
Tier by activity instead. Active policies and open claims belong on fast primary storage. In-force policies with no recent servicing sit on a middle tier. Settled, expired and closed records move to low-cost archival storage. And records past their retention period get purged under a documented policy rather than kept forever out of caution. Insurers routinely hold data long past any legal requirement because nobody wants to be the person who signed off on deletion. Understandable. Also expensive.
Object Storage for the Archive Layer
Why It Fits Insurance Better Than Traditional File Systems – Most insurance archive data is unstructured. Scanned proposal forms, PDFs, photographs, medical reports, video KYC recordings, call recordings from the servicing desk.
Object storage handles this far better than a traditional file system. Every document carries metadata alongside it, and that’s what makes retrieval practical. Instead of navigating a folder structure that made sense to whoever built it in 2011, you query by policy number, customer ID, product code, date range or document type.
Cloud object storage on AWS S3, Azure Blob or Google Cloud Storage, deployed in India regions to meet IRDAI data residency expectations, gives insurers affordable long-term storage with retrieval that doesn’t depend on institutional memory. On-premises object storage from NetApp, Dell or Pure Storage does the same job for insurers who’d rather keep the archive in their own environment.
Dealing With Legacy Data
Most Indian insurers are carrying data from at least two or three previous platforms. Some of it was migrated properly. Some were dumped into a file share during a rushed cutover. Some sits on tape that nobody has tested in years. A realistic approach starts with an inventory. What exists, where, in what format, and is it still readable? That last question catches people out more often than expected.
Â
Records still within retention and likely to be needed get migrated first with full metadata. Records within retention but rarely accessed get migrated with basic tagging. Records past retention get reviewed for deletion under a documented policy with legal sign-off.
Frequently Asked Questions
How long must Indian insurance companies retain policy records?
Retention periods vary by record type and product line under IRDAI guidelines, with policy records generally required through the contract term and for a defined period after settlement or termination. Insurers should confirm current requirements against applicable IRDAI circulars, as these have been revised over time.
Can insurers store archived policy records in the cloud?
Yes, provided the deployment keeps data within India-based cloud regions in line with IRDAI data residency expectations. AWS, Azure and Google Cloud all operate India regions suitable for this.
What’s the difference between backup and archive for insurance data?
Backup exists to restore recent data after a failure and is typically retained for weeks or months. Archive exists to preserve records for regulatory and legal purposes over years or decades, and is optimised for retrieval rather than rapid restoration.
How should insurers handle records stored on legacy tape systems?
Start by testing readability, since tape degrades and drive compatibility changes. Records still within retention should be migrated to modern storage with proper metadata. Anything past retention should be reviewed for disposal under a documented policy.
Where Brilyant Can Help
Archive projects in insurance rarely fail on technology. They fail on planning, metadata discipline and the messy reality of legacy data nobody fully understands anymore.
Brilyant works with insurers to assess what’s actually in the archive, design tiered storage that matches how records are genuinely used, and migrate legacy data with the metadata structure that makes retrieval work years later. We deploy across Dell, HPE, Lenovo, NetApp and Pure Storage, alongside India-region cloud storage on AWS, Azure and Google Cloud where the workload suits it.
If retrieving a fifteen-year-old policy file currently takes your team days rather than minutes, that’s the conversation worth having.
Talk to Brilyant’s infrastructure specialists about your archive and retrieval requirements. Get in touch.
