Prosus (PRX) investor relations material

Vision and Strategy for AI-Driven Commerce and Ecosystem Expansion

• Building a global AI-powered lifestyle and commerce ecosystem, leveraging unique data from billions of interactions across Latin America, India, and Europe.

• Focus on embedding AI at the core of operations, aiming for a trillion-dollar scale and empowering the everyday economy.

• Deep AI innovation and investment since 2018, with over 10,000 tech staff, including 1,000 in AI.

• Transitioning from loyalty-based engagement to life assistance and agentic operating systems, with loyalty programs increasing customer retention and cross-selling.

• Over 5 million partners, $115 billion in total payment volume, and $65 billion in gross merchandise value.

Large Commerce Models (LCM) and Technology Advancements

• LCMs are proprietary large language models trained on billions of real-world commerce interactions, powering intelligence for all applications and agents.

• LCMs enable deep personalization, intent recognition, predictive capabilities, and are more cost-effective than commercial LLMs.

• Already deployed at scale in over 40 applications, driving significant business impact and operational efficiency.

• LCMs are being rolled out across all group companies, with ecosystem-wide intelligence and further expansion planned through 2026.

• LCM for ecosystems and Toqan for Partners will launch in phases through 2026, starting in Latin America and Europe.

Agentic Workforce and Toqan Platform

• Toqan enables employees to build custom AI agents, democratizing automation and workflow optimization.

• Over 40,000 agents created, with 5,000 used daily across operations, HR, legal, sales, and engineering, delivering over 1,000 FTE equivalent impact.

• Agents automate complex tasks, saving significant time and enabling greater agility and independence.

• Toqan background tasks allow agents to proactively execute work triggered by real-world events.

• The platform is being extended to partners, allowing them to build and deploy their own agents.