Insurance Insight
A Power BI estate, replaced by an analyst that never sleeps.
- Client
- A pan-Asian life & takaful insurer (Malaysia)
- Replaces
- Power BI dashboards
- Stack
- Next.js 14DuckDB (in-memory)RechartsEntra ID SSOSynthetic data · no PII
Leadership ran the business on a sprawl of Power BI dashboards. Every new question meant a ticket to the BI team and a multi-week wait for a new report — while the numbers that mattered (persistency, loss ratio, agency productivity) sat one drill-down too deep to see in time.
- 1
Modelled the full insurance domain — New Business (APE/ANP), Portfolio in-force, Persistency (13/25-month), Claims (loss ratio, settlement), Distribution and a Shariah Takaful-fund view — over a fast in-memory DuckDB engine.
- 2
Made every section answer the question, not just chart it: ask in plain language, get the cut you need; the agent narrates what moved and why.
- 3
Built governed, one-click export and a finding-to-ticket action, with SSO, admin row-level security and version history baked in.
- 4
Trained and demoed entirely on synthetic, deterministic data — zero customer PII anywhere.
Static dashboards, refresh on a schedule
Ask in plain language, get the exact cut on demand
You read the chart and guess the cause
It narrates what moved and why (e.g. the cohort behind a persistency dip)
Export to Excel, then email it yourself
One-click governed export; raise a ticket straight from a finding
Per-seat licensing, every change gated by IT
Self-hostable, SSO, versioned, admin-controlled row-level security
This case study describes a prototype delivered on synthetic, deterministic data to demonstrate the approach. The client is anonymized and no customer PII is used anywhere.
Want the agentic version of your workflow?
Tell us what you're replacing. We'll usually answer with a demo.