Aegis: The Insurance Reserving Solution Built by Appointed Actuaries

IBNR reserving estimates liabilities for claims incurred but not reported, plus further development expected on known claims. It is typically the largest judgment-driven reserve on a general insurer's balance sheet. Aegis calculates IBNR by segment and accident period, with every method choice and override recorded.

Build development triangles, review loss development factors, project ultimate claims with one or more methods, then select final reserves by accident period. Aegis runs this as a guided wizard so the mechanics happen automatically and every judgment stays visible.

Triangles generate inside the platform instead of a spreadsheet chain. History loads once instead of being rebuilt every cycle, the SAMA template generates in one click instead of manual population, and every override is logged instead of living in someone's memory.

Yes. The guided wizard structures each step, from triangles to ultimate selection, so a junior actuary can produce a defensible first cut. Senior actuaries and the appointed actuary then review rather than rebuild.

Aegis outputs, full triangles, documented factors and claim-level drill-down, are built for external review. Prima's own actuaries also offer appointed actuary reserving support for method questions and regulator conversations, if wanted.

Data moves encrypted by secure upload or API, with role-based access separating claims entry, premium entry and actuarial review. Historical data is stored once, under version control, and never re-imported, and every change carries an audit record. Hosting and certifications are agreed during scoping.

One-time setup, loading history, defining segments and configuring validation, takes a few weeks. After that, each quarter adds only the new period.

Yes. Reinsurance runs on a net-to-gross ratio or separate triangles, and salvage and subrogation can be modeled as their own triangle, configured once at setup.

Yes. IBNR calculation for motor and medical insurance is the most common GCC configuration, alongside property, marine, engineering, health SME and takaful segments.

Claims reserving in general insurance arranges historical claims into development triangles, projects each accident or underwriting year to ultimate cost, and holds the gap above paid-plus-case as IBNR. Aegis automates the arrangement and projection.

Yes. Aegis runs as cloud actuarial reserving software, with data arriving by secure upload or API. Hosting region and on-premise options are agreed during scoping.

Yes, in one click per valuation, covering gross, salvage-adjusted and reinsurance-net triangles. Prima maintains the template logic in-house, so SAMA IBNR submission changes are absorbed in days.

Chain ladder, Bornhuetter-Ferguson and expected loss ratio ship today. Stochastic methods are on the roadmap, and triangles export cleanly to R and Python for teams running bootstrap or Mack analysis now.

Machine learning reserving covers AI-assisted method and factor selection. Aegis runs deterministic methods today with full actuarial control. AI-assisted selection that suggests weights and LDF exclusions is in development, with the actuary always deciding. Machine-led reserving is a direction, not a current claim.

IBNR covers claims that have occurred but have not yet been reported. IBNER covers further development expected on claims that are already known and reported. Aegis reports both as part of the same insurance reserving process.

Chain ladder projects purely from observed development, so it reacts fast but swings on immature accident years. Bornhuetter-Ferguson blends observed development with an expected loss ratio, which stabilizes early periods. Aegis runs both together, so you compare rather than commit upfront.

Not at Aegis. Reserving tools built by consulting firms sometimes are, which is a fair concern. Aegis is the platform Prima's own actuaries use on reserving engagements, so it has a dedicated product team and a roadmap reviewed every quarter, not a side project.

Aegis provides a one-page business case pack: current Excel-cycle cost against Aegis, built for the internal approval conversation. Tool selections often stall at budget approval rather than at technical fit. If you have already chosen a reserving tool once and watched the budget stall, this pack is built for whoever stalled it.

A French-language variant of the product interface is planned for French-speaking European markets, where searches for "logiciel actuariel reserving" show steady demand. Aegis already supports multi-region reporting for Solvency II environments in the meantime.

Three things. The close calendar slips, because finance waits on reserves that arrive in week three instead of week one. Audit hours and findings increase, because every figure needs manual reconstruction to defend. And senior actuaries spend their time assembling data instead of exercising judgment, while junior actuaries cannot produce a first cut, which leaves the whole process dependent on one or two people.

Every triangle cell opens to the underlying claims, with counts, averages and development periods. An auditor's question about a specific number is answered by clicking the cell it came from, not by reconstructing the calculation from scratch.

Segments can be defined at line of business, product, or payor level. Payor-level granularity was added at the direct request of our European partners, and the same structure works for GCC and Asian portfolios.

Aegis produces IBNR by segment and accident period in a structure that maps directly to the liability for incurred claims. Insurers using Delta, the IFRS Tech IFRS 17 engine, get a direct hand-off from Aegis reserving output into IFRS 17 measurement and disclosure, so the actuarial number and the accounting number come from the same data.

Bespoke agentic AI is a purpose-built AI agent scoped to one narrow task for a specific client, separate from the platform's built-in AI-assisted selection. An example already in use: an agent that monitors medical claims for new development patterns after a government portal integration changed notification speed. These are client-specific builds, not a general-purpose feature.

Fifteen minutes with a practicing actuary on a portfolio like yours, then an optional scoping call on segments, data and price. No obligation either way.