Artificial Intelligence and Machine Learning

Move beyond the pilot phase and into production

Implementing Artificial Intelligence and Machine Learning in BFSI means navigating legacy infrastructure, fragmented data, and regulatory complexity. Entelect brings experience rooted in domain knowledge to help navigating these realities. Whether it's conversational banking, agentic workflows, dynamic risk scoring, or MLOps that keeps models accurate over time, we bring the experience to take AI from proof-of-concept to production.

20%

Increase in cancer screening rates through Discovery’s Personal Health Pathways programme.

30,000+

Members completed structured exercise goals for the first time through the programme.

R200M

Estimated savings in six months through earlier detection and improved health management.

To achieve this, they needed a hyper-personalised recommendation engine that had access to multiple data sources from different business units, enabling a comprehensive view of each member’s wellbeing.

Entelect developed a data orchestration engine to power the Vitality Recommender. The engine integrated clinical, behavioural and lifestyle data, interprets key health assessments and clinical records and then identifies the three most relevant next-best actions for each member.

Each next-best action is paired with a personalised reward. The actions are surfaced in the Health App, while a complementary rules-based system validates completion by checking claims or clinical data before confirming the reward.

20% increase in cancer screening: The Personal Health Pathways programme has resulted in a 20% increase in cancer screening rates, supporting earlier diagnosis and treatment.

Increase active lifestyle in members: More than 30 000 members have completed structured exercise goals for the first time, thanks to the programme.

R200 Million savings: Discovery have indicated that this earlier detection and better management delivered an estimate R200 million in savings in 6 months.

3 people holding cellphones with the Discovery app displaying

What We Do

Our Artificial Intelligence and Machine Learning Capabilities

Intelligent Assistants and Chatbots

Conversational AI that handles customer queries, guides users through complex processes, and reduces pressure on service teams.

Agentic AI

Agentic systems can autonomously execute multi-step workflows, coordinate across enterprise systems, tools, and data sources, and orchestrate complex tasks with minimal human intervention – with configurable guardrails, oversight, and control mechanisms that ensure safe and reliable execution.

Retrieval-Augmented Generation (RAG)

Ground large language models in your own data and knowledge bases to deliver accurate, context-aware responses – without the hallucinations or generic outputs of off-the-shelf models.

Large Language Model (LLM) Integration

Deploy and orchestrate foundation models within your enterprise environment – customised and governed to meet your specific use cases, security requirements, compliance obligations, and operational guardrails.

Customer Sentiment Analysis

Automatically analyse customer feedback, reviews, and interactions to surface insights that inform product decisions, service improvements, and retention strategies.

Conversational Banking

Implement AI-powered conversational banking experiences across chat, voice, and digital channels. Intelligent assistants can securely handle customer queries, guide users through complex financial processes, support transactional interactions, and deliver personalised assistance in real time.

Hyper-Personalisation

Deliver individualised experiences across digital channels by using machine learning to understand customer behaviour, preferences, and intent.

Fraud Detection and Prevention

Identify suspicious patterns and anomalies across transactions and behaviour in real time, reducing financial exposure while minimising friction for legitimate customers.

Risk Assessment and Scoring

Replace manual, rule-based risk processes with dynamic ML models that assess credit, operational, or compliance risk more accurately and consistently at scale.

Dynamic Pricing

Adjust pricing in response to market conditions, customer segments, and demand signals – powered by models that optimise for margin, conversion, or competitive positioning.

Automated Claims Handling

Accelerate claims processing by automating document intake, classification, and decision-making, reducing handling time and improving the customer experience.

Computer Vision

Extract meaning from images and video – whether that's verifying identity documents, inspecting assets, automating quality control, or processing visual data at a scale humans can't match.

Optical Character Recognition (OCR)

Digitise and extract structured data from documents, forms, and images, eliminating manual data entry and accelerating downstream workflows.

MLOps and Model Lifecycle Management

Build, deploy, monitor, and retrain models at scale with a robust MLOps framework – ensuring your AI investments remain accurate and aligned to business outcomes over time.

Analytics Engineering

Build analytics engineering solutions that fit your architecture, your controls, and your business needs.

Learn more

Data Engineering and Platforms

Design and build modern, scalable data platforms - from data modelling that creates structured views to data integration.

Learn more

Quality Engineering

Embedding automated testing and quality gates so teams ship faster, with confidence.

Learn more

Core and Enterprise Systems

Modernising the core platforms that run the business — resilient, scalable, and built to evolve.

Learn more

Context

Momentum set out to create a proof-of-concept that could enhance client engagement by detecting emotional distress during contact centre interactions, enabling more empathetic and proactive client support.

Approach

Entelect created a call centre analytics solution leveraging Azure. The platform transcribes and analyses call data to detect distress signals and surface actionable insights, using existing infrastructure to minimise complexity.

Outcome

The proof-of-concept enabled accurate identification of distressed clients, and can improve client experience through timely follow-up, identify missed first-call resolutions and streamline customer service operations.

A person holding a tablet looking at data

Context

Coca-Cola Bottling Company aimed to help independent retailers accurately predict demand, improving ordering decisions and reducing stock inefficiencies.

Approach

Entelect developed an AI-powered Intelligent Ordering System that uses historical and real-time data to recommend optimal order quantities. The solution was piloted with local retailers and scaled across multiple African markets.

Outcome

The system improved order prediction accuracy, reduced stock-outs and excess inventory, and streamlined the ordering process. It also provided Coca-Cola Bottling Company with deeper insights into retail behaviour, strengthening supply chain planning and digital innovation capabilities.

A man working on a computer with headphones on

Context

MTN Group sought to strengthen oversight of international remittance transactions across its Mobile Money (MoMo) ecosystem.

Approach

Entelect helped create a machine learning model that analysed large volumes of transactional patterns and behaviour to accurately identify where an international transaction is either sent or received illegally.

Outcome

The solution created a repeatable mechanism for MTN Group to recover outstanding commission revenue and implement remediation measures to prevent recurrence.

Move beyond the pilot phase and into production

Vitality: Providing hyper-personalised health guidance with Vitality’s personal health pathways

Our Artificial Intelligence and Machine Learning Capabilities

Partners and Alliances

Related Expertise

How We've Delivered

Ready to move beyond the pilot phase? Let's talk.