Vitalis Group SA Business Plan › Section 9 › Technology, Data & Innovation

9.1 Technology Strategic Principles

Vitalis is being built technology-first, with the explicit objective
of a 30–40% structural cost-to-income advantage over legacy insurers.
The technology strategy is anchored on five principles:

• Cloud-native. Hosted on a primary South African
  cloud region (AWS Cape Town) with disaster recovery in Johannesburg;
  full residency compliance under POPIA.
• Microservices and APIs. Every business
  capability — quoting, underwriting, claims, payments, rewards — is built
  as an independent service exposed via secure API.
• Single customer view. A unified data layer
  ensures every customer interaction across products draws on the same
  identity, history and preferences.
• Real-time decisioning. Machine-learning models
  power pricing, fraud detection, lapse prediction and personalisation in
  real time.
• Buy-not-build for commodities. Vitalis builds
  proprietary technology only where it produces competitive advantage;
  commodity infrastructure (KYC, payments, document) is integrated from
  best-of-breed vendors.

9.2 Reference Architecture

The Vitalis reference architecture is layered into five logical
tiers: customer experience, distribution, business services, data and
integration, and infrastructure. Each tier is independently scalable and
replaceable, eliminating the monolithic dependencies that lock incumbent
carriers into multi-year transformation programmes.

9.3 Data Strategy and the Behavioural Data Moat

Data is the strategic asset of the Vitalis platform. Three categories
of data are captured, stored and analysed: (i) policy and transaction
data; (ii) behavioural data from wearables, telematics and digital
interactions; and (iii) externally enriched data including credit,
fraud, and aggregate health datasets. By Year 3, the Company expects to
be processing more than five billion daily behavioural data points and
to have accumulated a proprietary dataset that materially improves
pricing accuracy.

9.4 Artificial Intelligence and Machine Learning

Vitalis applies AI and machine learning across five operational
domains, each with explicit governance, model risk management and bias
controls:

• Underwriting and pricing. Gradient-boosted
  models supplement traditional generalised linear models (GLMs);
  explainability via SHAP values; fairness tested by demographic
  cohort.
• Fraud detection. Real-time claims-stage models
  flag anomalies; integration with industry fraud-share network
  (SAFPS).
• Lapse and retention. Survival-analysis models
  predict lapse risk 90 days in advance; outbound retention treatments
  tested via uplift modelling.
• Customer service. Generative-AI chatbot for
  tier-1 enquiries, escalating to human agents with full conversation
  context.
• Wellness coaching. Personalised, evidence-based
  wellness recommendations delivered in-app and via partner
  network.

9.5 Cybersecurity and Resilience

Cyber risk is treated as a first-order strategic risk. Vitalis will
obtain ISO 27001 certification within 18 months of launch and SOC 2 Type
II within 24 months. The cyber programme is owned by the Chief
Information Security Officer (CISO) reporting to the COO and with direct
access to the Audit and Risk Committee of the Board. Annual cyber
insurance limits of R250M are placed with Lloyds and continental
European markets.

9.6 Technology Investment

Total technology investment over the five-year plan period is R740M,
comprising R280M in initial platform build (capital-funded), R280M in
feature build-out (opex through the five years), and R180M in run-rate
infrastructure and software licences across the period. This investment
is the principal source of the Company’s long-term cost advantage.

Confidential — this business plan is provided to prospective investors and lenders for evaluation purposes only and may not be reproduced or distributed without the written consent of Vitalis Group South Africa (Pty) Ltd.