Insurance data teams: less upkeep, more data products
Less upkeep, more data products. Most of the week on insurance data teams goes to upkeep: broken feeds, reruns, access queues and documentation. Data Workers hands that work to agents so the team ships the data products the business has been asking for.
Key takeaways
- •Who this is for. Data leaders and teams at carriers, MGAs and reinsurers.
- •The problem today. Most of the data team's week goes to monthly bordereaux mapping, reconciliation at close and stewardship spreadsheets.
- •What changes. Specialised agents take that work over across the systems you already run: they catch bad data where it enters, fix it where it lives and verify the number before anyone reads it.
- •What stays with your people. Anything regulated or customer-facing waits for a named approver with the full diff, blast radius and rollback plan in front of them.
- •The result. Data teams that spend their time on pricing, claims and distribution products.
The PRD for agentic data in insurance
This is the requirements document we would write with you. Everything below expands on it.
| Who | Data leaders and teams at carriers, MGAs and reinsurers. |
| The job | Underwriting, claims and distribution data that runs itself. |
| Environment | Snowflake, Databricks or BigQuery; policy admin and claims systems (Guidewire, Duck Creek); partner portals and SFTP; geospatial and cat vendors; dbt; Airflow, and the other platforms your teams run. |
| Constraints | Regulated and customer-facing numbers always go to a named approver; agents run in your infrastructure under the roles you already govern. |
| Success is measured by | Access for a new actuary: from about a week to 5 min; Finding the right dataset: from about 4 hours to 30 s; Bordereau layout change: from about 4 days to 15-30 min; catalog and ownership upkeep closed autonomously 86% of the time. |
| Owned by | The head of data or CDO, with domain owners as named approvers. |
| Pilot scope | Bordereaux tie-out at L2. |
What going fully agentic and autonomous means in insurance
- •Put your data pipelines on autopilot. Agents catch bad data where it enters, fix it where it lives and verify the number before anyone reads it.
- •Run the data back office without the busywork. The week that goes to monthly bordereaux mapping, reconciliation at close and stewardship spreadsheets moves to agents.
- •Scale data without scaling headcount. New sources, partners and data products are onboarded as missions, not projects.
- •Get the plumbing off your plate. Data teams spend their time on pricing, claims and distribution products.
The shift
Insurance is a data business that has always run on people. Actuaries reconcile, claims analysts rebuild extracts, distribution teams chase broker data, and data offices maintain stewardship spreadsheets. Most carriers now have a global data office and domain stewards, but the stewards spend their week tracking definitions by hand while the domains drift.
Meanwhile the data sources keep multiplying: delegated-authority bordereaux, catastrophe and geospatial enrichment, telematics, third-party data and a broker channel that sends everything in its own format.
Our view: the data office should set the standards and the agents should keep them true. Agents land and map partner data, build domain data products, verify premium, claims and exposure numbers, and keep definitions and ownership current in every domain. Actuaries and underwriters get data they trust on day one of the month, not day ten.
How data work in insurance got here
The policy-admin era. Each line of business had its own system and its own reports. Monthly close meant weeks of reconciliation.
The data office era. Carriers built enterprise data offices, data warehouses and stewardship programs, and started to treat data as products. The structure is right; the upkeep still runs on spreadsheets and tickets.
The agentic back office. Agents do the stewardship work and the engineering work: mapping, quality, documentation and access. The data office points the fleet at its standards.
The three eras side by side:
Each era changed the tools. Only the agentic back office changes who does the work.
A day on an insurance data team today
On the fifth business day of the month, an MGA's bordereau arrives with a new column order and two renamed fields. Written premium for that program loads short. The actuarial team sees the dip in the monthly pack on day eight and asks claims and finance which number is right. A data engineer finds the layout change on day nine. The pack goes out late, with a footnote.
That morning is typical, not exceptional. Across a week, this is where the team's time goes, today and with agents on the back office:
Time spent on new data products moves from 17% of the week to 50%. Most of the rest becomes review: people steer and approve agent work instead of doing it by hand.
Then and now
| Work | Today | With Data Workers |
|---|---|---|
| Bordereaux from MGAs and partners | Mapped by hand every month | Changes caught on arrival, mapped, tied to the partner's totals |
| Domain standards in the data office | Tracked in spreadsheets, drifting | Enforced by agents as org standards, gaps shown per domain |
| Claims data for actuarial and finance | Different versions in different places | One governed version, verified before use |
| New rating factors | Quarters to reach production | Built and tested as a mission, approved by a person |
| Telematics and third-party data | Consent and contract terms tracked outside the data | Enforced on every table and delivery |
| Broker and agent distribution data | Chased by email, reconciled at month end | Landed, mapped and tied to production and commissions daily |
These are the impact metrics we move:
- •Access for a new actuary: from about a week to 5 min
- •Finding the right dataset: from about 4 hours to 30 s
- •Bordereau layout change: from about 4 days to 15-30 min
Across the routine work:
The biggest change is Access for a new actuary: from about a week to 5 min. Work that waited in queues finishes in minutes or hours, because agents detect, fix and verify in one pass.
Use cases, in depth
1. The bordereau that changed shape
Today. Found when the monthly pack looks wrong. Fixed after the deadline.
With Data Workers. Ingestion flags the layout change on arrival, Schema Evolution maps the renamed fields, Pipeline Building updates the premium model, Quality Monitoring ties written premium to the MGA's control totals, and Conductor holds the actuarial refresh for one approval.
What changes. The monthly pack goes out on time, without a footnote.
2. Making the data office's standards real
Today. The office defines a standard (every certified data product has an owner, an SLA and a governed definition) and stewards check it by hand once a quarter.
With Data Workers. Autopilot turns the standard into an org rule. Data Context & Catalog finds every data product in scope, proposes owners from usage history, Context Wizard reconciles conflicting definitions and flags the ones that need a steward, and the dashboard shows each domain's gap closing week by week.
What changes. Standards stop drifting, and stewards spend their time on decisions rather than spreadsheets.
3. A new rating factor from geospatial and catastrophe data
Today. Pricing wants a new factor from a geospatial enrichment vendor. Sourcing, joining, testing and productionizing it takes a quarter.
With Data Workers. The factor is built as a mission: Connectors and Ingestion land the vendor data, Pipeline Building joins it to the policy model with tests, Quality Monitoring checks coverage and outliers, and the change to pricing inputs waits at L2 for the pricing actuary.
What changes. Pricing teams test ideas in weeks.
4. Distribution data through enrollment season
Today. For carriers selling through brokers and agents, enrollment periods bring volume spikes, late partner files and commission disputes.
With Data Workers. Connectors land carrier, enrollment and broker feeds; Quality Monitoring ties enrollments to production and commissions daily; Conductor keeps reporting fresh through the peak. See the sales data mart playbook.
What changes. Distribution leaders see production by channel every morning, and commission disputes start from the same data.
Agents working together: A bordereau that changed shape mid-month
The bordereau from the first use case, end to end. Five agents and one approver correct written premium before the monthly pack is built.
Minute by minute
This is an illustrative example. Bordereaux land in an S3 bucket and load into Snowflake; dbt builds premium and claims marts read by actuarial on day 7.
| Time | Who | What happens |
|---|---|---|
| Day 5 09:00 | Coverholder | Bordereau uploaded with reordered columns |
| 09:02 | Schema Evolution agent | detect_schema_change flags the layout |
| 09:06 | Quality Monitoring agent | Mapping proposed; run_quality_check against control totals |
| 09:12 | Pipeline Building agent | Model diff with tests; ticket opened |
| 10:40 | Spellbook Data Catalog | Actuarial lead approves |
| 10:45 | Autonomous Data-Conductor, via Airflow | Rebuild; premium ties |
How it fits your stack
Data Workers connects to what you already run: Snowflake, Databricks or BigQuery; policy admin and claims systems (Guidewire, Duck Creek); partner portals and SFTP; geospatial and cat vendors; dbt; Airflow. Each agent is an MCP server, so the same agents work inside Claude Code, Cursor or Codex and in Slack or Teams through Spellbook. Nothing migrates.
- •Runs in your infrastructure with your credentials and your model key. The hosted Autonomous Data-Conductor sees workflow metadata only, never rows.
- •Uses the access you already govern. Agents act under scoped service identities and your existing roles; every grant goes through provision_access and is logged.
- •Proposes changes as diffs in your dbt project, orchestrator and warehouse, with blast radius, tests and a rollback point, and applies them only at the autonomy level you set for that domain.
- •Open-source core. The Apache-2.0 community edition ships 11 agents and 160+ MCP tools, so your engineers can read exactly what the agents call.
How autonomy expands over the first year
Autonomy is earned per task class. In a mature deployment, it looks like this:
Catalog and ownership upkeep closes autonomously 86% of the time. For Pricing inputs and reserves, autonomous closure is 0% by design: every change is proposed with a full diff and approved by a named person.
A typical path to get there:
- •Month 1, L1. Agents map every domain, data product, definition and partner feed.
- •Months 2 to 3. Catalog upkeep, cleanup and routine access at L4. Bordereaux mapping in staging, reruns and quality rules at L3.
- •Months 4 to 6. Org standards enforced across domains. New rating factors and data products built as missions.
- •Months 6 to 12. Partner mappings with a clean record move up. Pricing inputs, reserves and financial-reporting numbers stay at L2. Underwriting and claims decisions stay with people.
The same path as a curve, month by month:
By month 12, more than 38% of agent actions run fully autonomously and manual work falls below 10%, while pricing and reserve changes stay with a named approver.
What happens when an agent is wrong
- •At L2 propose, nothing changes until a named person approves. The proposal shows the full diff, the blast radius and how to roll it back. Unanswered requests expire and escalate.
- •At L3 act, reversibly, every change keeps a rollback point, and Conductor verifies the downstream number after the change. If verification fails, the change is reversed and the domain drops a level.
- •No agent approves or promotes its own work. Autonomy only grows on evidence, per repair class, and repeated errors shrink scope automatically.
- •One org-wide stop halts all autonomous dispatch, and every action lands in a hash-chained audit trail you can export with generate_audit_report.
Governance and regulation
Insurers answer to regulators and auditors for reserves, pricing and financial reporting, and telematics data needs a consent basis. Data Workers provides lineage, named approvals and an audit trail, and enforces consent on deliveries.
Data Workers doesn't make you compliant; your controls, people and assessors do. It supplies controls and evidence wherever agents touch data. Not legal advice.
The alternatives insurance data teams weigh
Build it yourselves. A strong team can wire a coding agent to MCP servers for the warehouse and dbt. The work is everything around it: routing approvals to the right owner, blast radius, rollback, receipts, and current context about every other system.
Observability, catalogs and point tools. Observability alerts, catalogs document, and quality tools test. Data Workers takes their signals, diagnoses, proposes the fix, verifies it and leaves the receipt.
Platform-native agents. Snowflake made its managed MCP server and Cortex Agents generally available on Nov 4, 2025, and Databricks made Genie One generally available on Jun 16, 2026. They are excellent inside their platform; this work crosses the source systems, orchestrator, warehouse and BI, and Data Workers works across all of them.
In this space specifically. Policy, claims and bordereaux tools each own a step; Data Workers keeps the data across them tied out and governed.
The case for your CTO
What this replaces. The week spent on monthly bordereaux mapping, reconciliation at close and stewardship spreadsheets.
What your teams get back. Time for pricing, claims and distribution products, on data that is verified before anyone reads it.
Integration effort. Connecting is configuration, not a migration. The agents attach over MCP to the platforms you already run, under the roles you already govern, inside your own infrastructure.
Risk and control. Autonomy is set per domain, starts at L1 and rises only on evidence. Regulated and customer-facing numbers stay at L2 with a named approver, and one org-wide stop halts all autonomous work.
Why now. Delegated authority and enrichment multiply sources.
The first win. Bordereaux tie-out at L2. The pilot below lays out the scope, what you see each week and the cost.
The climb: from assisted to autonomous to agentic
Our thesis is that the data platform climbs the way analytics did, in stages. In insurance, that climb looks like this:
- •Assisted. Agents work alongside each data team, read the whole stack and propose every fix with a receipt. Regulated and customer-facing numbers always go to a named approver.
- •Autonomous, under your governance. The upkeep behind monthly bordereaux mapping, reconciliation at close and stewardship spreadsheets runs on its own, domain by domain, as the receipts earn it.
- •The agentic enterprise. The platform largely runs itself. People set direction and standards, and the teams spend their time on pricing, claims and distribution products.
You choose the altitude, every step is visible and reversible, and the pilot below is the first rung.
The pilot: how insurance data teams start
The pilot is how a team proves the value on its own data before committing.
- •What we connect. Snowflake, Databricks or BigQuery; policy admin and claims systems (Guidewire, Duck Creek); partner portals and SFTP; geospatial and cat vendors; dbt; Airflow, over MCP, inside your infrastructure. Nothing migrates.
- •Scope. Bordereaux tie-out at L2.
- •First week. Agents run at L1 and show the fixes they would have made, with receipts.
- •By week two. Reversible upkeep (reruns, staging fixes, documentation, cleanup) runs on its own; regulated and customer-facing numbers stay at L2 with a named approver.
- •At day 30. We review the impact metrics against your baseline:
- •Access for a new actuary: from about a week to 5 min
- •Finding the right dataset: from about 4 hours to 30 s
- •Bordereau layout change: from about 4 days to 15-30 min
- •the share of catalog and ownership upkeep closed autonomously, targeting 86%
- •What you decide. Which domain or team goes next. The pilot fee is credited in full against the first year.
- •Cost. Start with a pilot ($7,500 one-time, credited in full against the first year). Scale is from $1,000 a month and Enterprise from $3,000 a month, billed annually, with unlimited seats, no usage meter and no markup on model spend. See pricing.
Teams we work with in insurance
FAQ
Guidewire? Through the warehouse and APIs.
Reserves? L2, decisions manual.
Telematics consent? Enforced per delivery.
Where do the agents run? In your infrastructure, with your credentials and model key. The hosted Conductor sees workflow metadata only.
Do we have to migrate anything? No. Your warehouse, dbt project, orchestrator, catalog and BI stay where they are.
Sources
- •Chubb, Head of Data Products: https://www.dreamworkhq.com/job/a310ec2e-591e-4031-af1a-8a200e3aba71
- •Data Workers pricing: https://dataworkers.io/pricing/
- •Data Workers open-source repository (tool registrations): https://github.com/DataWorkersProject/dataworkers-claw-community
- •Snowflake, managed MCP server and Cortex Agents GA release note (Nov 4, 2025): https://docs.snowflake.com/en/release-notes/2025/other/2025-11-04-cortex-agents-mcp
- •Databricks, Genie One launch (Jun 16, 2026): https://www.databricks.com/company/newsroom/press-releases/databricks-launches-genie-one-all-new-agentic-coworker-every-team