Data Scientist

Vodafone; Midrand; On-site; ZAR 900,000 - 1,500,000; Full time

Job summary

Vodacom is seeking a Data Scientist specializing in Generative AI and Agentic AI to design and deliver next-generation, AI-powered customer experience solutions. You will build production-grade LLM-powered systems and agentic workflows enabling real-time decisioning, intelligent automation, and proactive, personalized engagement across Vodacom’s digital channels and markets.

This role operates at the intersection of LLMs, agent orchestration, and customer intelligence, transforming raw customer

Qualifications

• Experience designing and delivering GenAI/LLM powered CX solutions.
• Proficiency with AI architectures, prompt engineering, and model eval.
• Knowledge of data pipelines and MLOps for production use.

Responsibilities

• Design, build, and deploy LLM-powered applications for customer experience.
• Develop RAG architectures and agentic AI workflows for CX journeys.
• Implement monitoring, safety controls, and production-readiness practices.
• Enable real-time decisioning and proactive customer engagement across channels.

Skills

Generative AI

LLM systems

Agentic AI

RAG

Job description

When it comes to putting people first, we're number 1.

The number 1 Top Employer in South Africa.
Certified by the Top Employer Institute 2026.

Vodacom is a leading African mobile communications company, serving over 200 million customers across South Africa, Tanzania, DRC, Mozambique, Kenya, Lesotho, Egypt, and Ethiopia. Our mobile networks reach a population of approximately 250 million people. Through Vodacom Business Africa (VBA), we also deliver managed services to enterprises in over 40 countries across the continent.

We are truly data-driven, committed to delivering intelligent solutions that enhance customer experience and business performance. Our Big Data and AI team plays a pivotal role in enabling this transformation by supporting the CX and Digital space with scalable, innovative data capabilities.

Role Purpose/Business Unit:

We are looking for a Data Scientist specializing in Generative AI and Agentic AI systems to design and deliver next-generation, AI-powered customer experience solutions.

This role is focused on building production-grade LLM-powered systems and agentic workflows that enable:

• Real-time decisioning
• Intelligent automation
• Proactive and personalized customer engagement

You will operate at the intersection of LLMs, agent orchestration, and customer intelligence, delivering scalable solutions across Vodacom’s digital channels, customer care platforms, and markets.

This role operates across text, voice, and multimodal customer data, transforming raw customer interactions into intelligent, AI-driven actions at scale.

Your responsibilities will include:

GenAI & LLM System Development (Primary Focus)

• Design, build, and deploy LLM-powered applications including:
  • Retrieval-Augmented Generation (RAG)
  • Conversational AI
  • Summarisation, classification, and recommendation systems
• Develop RAG architectures integrating structured and unstructured enterprise data
• Implement robust prompt engineering, evaluation frameworks, and guardrails
• Build LLMOps pipelines covering orchestration, monitoring, evaluation, and optimisation
• Ensure solutions are scalable, secure, and production-ready

Agentic AI & Workflow Automation (Core Capability)

• Design and implement agentic AI systems capable of:
  • Multi-step reasoning and planning
  • Tool and API orchestration
  • Autonomous execution with feedback loops
• Develop multi-agent workflows to support:
  • Customer query resolution
  • CX insights generation
  • End-to-end journey orchestration
• Implement human-in-the-loop mechanisms, approvals, and safety controls
• Integrate AI agents into:
  • Chatbots and virtual assistants
  • IVR and voice systems
  • Backend operational workflows
  • Relevant application workflows

Customer Experience (CX) Intelligence (High Impact)

• Design and build scalable AI solutions to extract value from complex unstructured customer data, including:
  • Call centre audio recordings and voice data
  • Speech-to-text transcripts and conversational logs
  • Chatbot and digital interaction data
  • NPS and survey verbatims (free-text feedback)
  • Customer emails, service requests, and support tickets

• Develop end-to-end pipelines that transform raw unstructured data into actionable intelligence using:
  • LLMs and Generative AI
  • NLP and speech/voice analytics
  • Multilingual processing techniques

• Build models and LLM-driven systems to enable:
  • Sentiment, emotion, and behavioural signal detection (text + voice)
  • Customer intent classification and journey mapping
  • Root cause analysis and large-scale theme extraction
  • Call summarisation, tagging, and quality evaluation
  • Identification of churn signals, friction points, and experience drivers

• Deliver real-time and near real-time CX intelligence, enabling:
  • Dynamic next-best-action recommendations
  • Proactive issue detection