From implementation to optimisation, orchestration and AI-enabled actuarial operations.
Introduction: What Comes After Implementation?
For much of the last decade, actuarial modernisation has been driven by technology transformation. Insurers have invested in new actuarial platforms, implemented IFRS 17, replaced legacy systems and introduced more sophisticated modelling environments.
These initiatives were necessary. They established the technology, models and reporting capabilities required to support increasingly complex actuarial functions.
But for many insurers, the next challenge is no longer simply getting technology in place. It is making those capabilities work together effectively.
A modern technology environment does not automatically create a modern actuarial operating model. Even with sophisticated platforms, actuarial functions can remain dependent on manual processes, fragmented data, disconnected models, repeated reconciliations and significant human intervention.
This raises a more important question for actuarial leaders: How effectively does the entire actuarial operating model work? The next phase of actuarial modernisation is about moving beyond implementation towards operational optimisation, end-to-end workflow orchestration and, ultimately, AI-enabled actuarial operations.
Phase 1: Modernisation Through Technology Implementation
The first phase of modernisation focused on establishing the necessary technology and compliance foundations.
For many insurers, the priorities included:
- Digitalising actuarial processes.
- Implementing IFRS 17.
- Replacing or upgrading legacy technology.
- Introducing modern modelling platforms.
- Improving reporting capabilities.
- Establishing new data structures.
The emphasis was on implementation: introducing the systems and capabilities needed to meet regulatory requirements and support the business.
For many organisations, this represented a substantial transformation. However, completing an implementation does not necessarily mean that the wider operating model has become efficient.
A system can perform exactly as intended while the processes surrounding it remain fragmented. Data may still require manual transformation, model outputs may need to be transferred between systems, and teams may continue to spend considerable time reconciling results.
This is the distinction between implementing technology and modernising the way the actuarial function operates.
Implementation asks: Does the technology work?
Modernisation asks: Does the operating model work?
The next phase begins when the focus shifts from what has been implemented to how effectively those capabilities are used.
Phase 2: Modernisation Through Operational Optimisation
The second phase is about improving the performance of the entire actuarial operating model.
Rather than focusing solely on individual platforms or processes, the objective is to reduce operational friction, improve connectivity and create more efficient end-to-end workflows.
Four priorities underpin this progression.
1. Optimise
Identify and remove unnecessary steps, duplication and inefficiencies across the actuarial lifecycle.
Before automating a process, organisations need to understand whether the process itself is effective. Automating an inefficient workflow can simply make an inefficient process run faster.
2. Automate
Reduce repetitive manual activity and improve consistency through automation.
This can include routine calculations, data preparation, recurring processing activities and other tasks that follow established rules.
Automation creates value, but its impact is limited when the wider workflow remains disconnected.
3. Integrate
Connect data, models, systems and downstream processes so that information can move more effectively across the actuarial function.
Integration reduces the need for repeated data transfers, manual reconciliations and duplicated transformation activities. It also helps create a more consistent flow of information between actuarial, finance and reporting teams.
4. Orchestrate
Move beyond automating individual tasks towards coordinating the end-to-end actuarial workflow.
An orchestrated process connects the activities that make up the actuarial lifecycle, from data receipt and validation through model execution, output analysis, exception management, human review and reporting.
The objective is not simply to make individual tasks faster. It is to make the entire operating model work more effectively.
This progression can be summarised as:
Optimise → Automate → Integrate → Orchestrate
The distinction matters. Automating a single process may save minutes or hours. Orchestrating the end-to-end process can change how the actuarial function operates.
The Hidden Cost of Fragmentation
One of the most persistent challenges in actuarial modernisation is fragmentation between otherwise capable systems, models and processes.
Consider a typical actuarial workflow.
Data originates in one system and is transformed through another process. A model consumes that data and produces outputs. Those outputs are exported, checked and reconciled. Finance receives results, reporting teams assemble information, and exceptions are identified for further investigation.
Finally, the actuary reviews the information and applies professional judgement.
Each individual component may work effectively. Yet the overall workflow can still be inefficient.
Fragmentation between data, models, finance, reporting, the general ledger (GL), scenario analysis and governance can create several operational challenges:
- Repeated hand-offs:
Information moves between teams and systems, requiring additional coordination. - Manual reconciliation:
Teams spend time checking and resolving differences between outputs. - Duplicated transformation:
Data is processed more than once across disconnected workflows. - Disconnected controls:
Validation and governance activities may be spread across different stages of the process. - Delayed insight:
Information takes longer to move through the lifecycle and reach decision-makers. - Reduced actuarial capacity:
Skilled professionals spend more time on production and repetitive execution, leaving less capacity for interpretation and analysis.
These costs are not always visible when individual systems are assessed in isolation. A model may run efficiently, a reporting platform may function correctly, and a data process may meet its requirements, while the hand-offs between them continue to consume time and resources.
This is why modernisation cannot stop at improving individual components.
The opportunity lies in reducing friction between those components and improving how the whole actuarial workflow operates.
Addressing fragmentation creates the conditions for greater consistency, more effective use of existing technology and a stronger foundation for intelligent automation.
Automation Isn’t the End Goal
Automation is an important part of actuarial modernisation, but it should not be mistaken for the final objective.
Automating individual tasks can improve efficiency within a specific process. However, if those tasks remain disconnected, the wider operating model may still depend on manual intervention and repeated coordination.
The progression is therefore broader than task automation alone:
Manual process → Automated task → Integrated workflow → Orchestrated operating model
Consider a workflow in which data arrives, undergoes validation, feeds into model runs, produces outputs for checking, and then moves into reporting and exception management.
Automating the model run may improve one stage of the process. But if data validation remains manual, outputs need to be transferred between systems, and exceptions must be identified through separate activities, significant operational friction remains.
An integrated and orchestrated workflow aims to coordinate these activities more effectively.
This requires organisations to look beyond the performance of individual tasks and examine how information, processes and decisions connect across the actuarial lifecycle.
The goal is not automation for its own sake. It is an operating model in which technology supports a more coordinated, repeatable and efficient end-to-end process.
That distinction becomes increasingly important as insurers consider the next stage of modernisation: AI-enabled actuarial operations.
Phase 3: The AI-Enabled Actuarial Function
As actuarial operating models become better integrated and orchestrated, AI and intelligent automation create opportunities to change how work is executed and where actuarial expertise is applied.
The emerging possibility is a workflow in which technology does more than execute predefined instructions. It can increasingly support the coordination of activities across a process, help identify exceptions and direct attention towards areas requiring human review.
A future-oriented workflow could look like this:
Data → Models → Validation → Exception management → Actuary → Decision
AI & automation increasingly support execution, coordination and workflow across the process.
Within this model, technology can increasingly support activities such as:
- Execution of routine processes.
- Coordination of workflow activities.
- Routine validation.
- Identification of exceptions.
- Workflow management.
Actuarial professionals can then focus more of their capacity on the areas where their expertise is most valuable:
- Professional judgement.
- Interpretation of results.
- Challenging assumptions and outcomes.
- Governance and oversight.
- Communication of insights.
- Decision support.
The intention is not to remove the actuary from the process. It is to reduce the operational burden surrounding actuarial judgement, allowing professionals to concentrate on interpretation, challenge and decisions.
AI does not remove actuarial expertise. It changes where that expertise creates value.
However, this transition depends on the quality of the operating foundation beneath it.
Why AI Isn’t the Starting Point
AI-enabled operations are an emerging stage of actuarial modernisation, but implementing AI alone does not resolve the underlying challenges of an actuarial function.
If data is fragmented, AI inherits that fragmentation. If processes are inefficient, automation can accelerate inefficient processes. If models are poorly integrated, AI does not automatically resolve the integration problem. And if governance is unclear, introducing AI creates additional questions rather than removing them.
The foundation matters.
Organisations need to understand their processes, improve how systems and models connect, establish appropriate controls and create a more coherent operating model before expecting intelligent automation to deliver its full potential.
This is not an argument against AI. It is a recognition that its effectiveness depends on the environment in which it operates.
The six enablers of actuarial modernisation provide a useful framework for assessing that foundation.
The Six APMTM Enablers
At MBE Consulting, the APM™ Framework considers actuarial performance across six interconnected enablers: People, Processes, Methodologies, Systems, Models and Data. Together, they provide a useful lens for assessing the operating foundation required for modernisation
PEOPLE
Do teams have the skills and capacity required for a modern actuarial function?
Modernisation must consider how people work alongside technology and how actuarial expertise is applied as routine activities become increasingly automated.
PROCESSES
Are workflows efficient, repeatable and scalable?
Processes should minimise unnecessary steps, duplication and manual hand-offs while supporting consistent execution across the actuarial lifecycle.
METHODOLOGIES
Are actuarial methodologies designed for the current business and reporting environment?
Methodologies need to support the requirements of the organisation and work effectively within its broader modelling, reporting and operational processes.
SYSTEMS
Do technology platforms support the required operating model?
Systems should enable the necessary connectivity between activities rather than simply functioning as separate solutions.
MODELS
Are models efficient, maintainable, transparent and fit for purpose?
Models remain central to actuarial work, but their effectiveness also depends on how they connect with data, systems and downstream processes.
DATA
Is data accessible, controlled, consistent and usable?
Data underpins the actuarial lifecycle. Its quality, accessibility and consistency influence the effectiveness of modelling, reporting, validation and any subsequent AI-enabled processes.
Technology touches all six enablers, but technology alone does not resolve every challenge within them. Together, these six areas provide the foundation for operational optimisation and the intelligent automation that may follow.
Bringing the Three Phases Together
The evolution of actuarial modernisation can be understood as three connected phases.
Phase 1: Technology Implementation
Digitalise processes, implement IFRS 17, replace legacy technology and establish the necessary platforms and capabilities.
Phase 2: Operational Optimisation
Optimise processes, automate repetitive tasks, integrate data and systems, and orchestrate end-to-end workflows.
Phase 3: Intelligent Automation
Build towards AI-enabled actuarial operations in which technology increasingly supports execution, coordination and exception identification, while human expertise remains central to judgement and decision-making.
The six enablers (People, Processes, Methodologies, Systems, Models and Data) provide the foundation throughout all three phases. Automation and AI are capabilities that can be layered across the operating model as that foundation develops.
The phases are connected rather than isolated. The progress made in implementation creates the opportunity for optimisation, while a stronger and more integrated operating model creates better conditions for intelligent automation.
The objective is to move from having modern technology to operating a modern actuarial function.
The Question Actuarial Leaders Should Now Be Asking
The next generation of actuarial modernisation is not simply about buying another platform. It is not simply about implementing AI, and it is not finished when IFRS 17 goes live.
The more important question is:
How effectively does your actuarial operating model convert data and models into timely, reliable and decision-useful insight?
Answering this requires a broader view of the actuarial function: how people work, how processes operate, how methodologies are applied, how systems connect, how models perform and how data moves across the organisation.
It also requires recognising that efficiency is determined not only by the performance of individual components, but by how effectively the entire workflow operates.
For insurers that have already invested significantly in actuarial technology, this represents an important opportunity to realise more value from their existing capabilities.
The next phase begins with optimisation, progresses through integration and orchestration, and creates the foundation for AI-enabled actuarial operations.
Ultimately, the goal is not simply to modernise technology. It is to modernise how the actuarial function delivers value.
