Meridian's core systems are now the constraint on growth, not the balance sheet.
Underwriting talent is strong. Capital is strong. But a 34-year-old policy administration system caps how fast Meridian can price, bind, and pay — and every competitor built after 2015 already prices in seconds.
This is not a technology refresh. It is a three-year capital allocation decision: where to modernize, where to partner, and where to retire complexity outright.
What follows is the diagnosis, the three paths we evaluated, the recommended sequence, and the specific decisions we need from this steering committee before quarter-end.
The policy administration system was built for paper renewals; every digital feature since has been bolted on, not designed in.
Actuarial changes queue behind a single mainframe release train shared with claims and billing.
Manual triage and rekeyed data drive an 18% error rate on first-notice-of-loss, feeding loss-ratio drift.
Two senior underwriting leads left this year citing tooling, not comp — the system is now a retention risk.
Replatform policy administration onto a modern core over 30 months. Highest capability upside, highest execution risk.
License a modern core-as-a-service and integrate via API, preserving underwriting IP while renting the plumbing.
Retire the legacy core into a managed run-off and migrate only active books, freeing budget for growth lines.
Meridian doesn't need to build a policy admin system from scratch to compete on speed. License a proven core-as-a-service for the two growth lines now, and reserve a full core replatform for year three, once the operating model has proven it can absorb change.
We've debated a core replacement for six years. This is the first plan where I can tell the board exactly what changes in year one.
Each priority ships a working capability before the next one starts — no big-bang cutover, no 18-month blackout.
API gateway between the legacy core and the new partner platform, live by week six.
Route small-commercial quoting through the new core; legacy stays system of record until parity is proven.
Deploy FNOL triage automation in claims, targeting a 30% cut in rekeying errors this quarter.
A joint Meridian–vendor steering committee owns scope; the CIO chairs, and the transformation office runs the cadence.
Partner licensing plus integration costs roughly $6M in year one, funded by claims-leakage savings starting month nine.
A full core replatform — the year-three decision gate — adds $12M, contingent on the pilot's proof points.
Claims-leakage recovery and faster quote-to-bind start paying back before the core-replatform decision gate, de-risking the year-three commitment. (Curve shapes are illustrative.)
Root-cause the core constraint and quantify leakage. Complete — this deck.
Target operating model, API contracts, and the investment case, board-approved.
One growth line live on the partner core; legacy stays system of record.
Expand to remaining commercial lines; begin core-replatform discovery.
The new core is system of record; the change cadence becomes business as usual.
We need three decisions from this committee by Friday: fund the $6M year-one partner integration, name the pilot line owner, and set the month-22 breakeven review as a standing board item. Everything else in this roadmap follows from those three signatures.