Ioannis Zempekakis

ToqanClaw: Everyone Can Become A Builder

The core message of the session was that ToqanClaw is intended to let non-engineers become builders by creating AI applications without installations, updates, or code. Ioannis explained that the team took the main idea of OpenClaw and rebuilt it bottom up inside their proprietary, enterprise-ready platform so it would be safe, secure, and usable by everyone. He emphasized that users can simply type what they want, see results, and build applications on top of ToqanClaw. He also framed the launch as a European product release at scale, saying the team is excited that the technology was made in Europe. The session then moved from the value proposition to concrete use cases. Ioannis described lawyers, accountants, finance people, engineers, marketers, and HR teams producing thousands of solutions once a powerful tool is placed in their hands. He said building an application can take literally two minutes, as can building and sharing a presentation, while noting that teams are increasingly using HTML and websites rather than presentations because they are more live and powerful for communicating information. He also described real-time dashboards, traffic and weather monitoring, logistics optimization, and integrations with Slack or Whatsapp that can send action-oriented briefings connected to Prosus context, data, and insights. In the live demo, Ioannis showed a new chat workflow for generating an HTML presentation or application, installing a ToqanClaw guide skill from an internal marketplace, and sharing the resulting app with named colleagues or at an organization level. He highlighted that ToqanClaw is “blazing fast” because it uses open source models as well as closed source models, selecting the best model per task. He then shared adoption figures inside Prosus: 60,000 agents created with Toqan, 20,000 actively used and delivering value month over month, and 13,000 applications already built and connected to intelligence, the LCM, and company data. Tara from Legal demonstrated how a non-engineer can use ToqanClaw for practical business workflows. She built an online brand protection monitor for food-related brands and partners across mobile app stores and other platforms, then showed a restaurant monitoring application using the Just Eat API to detect low ratings, delivery-time issues, and franchise-level problems. Ioannis used her example to underline how powerful tools plus curiosity can create unexpected use cases. The session closed with an LCM for partners demo for iFood restaurants, showing how an agent could identify customer clusters, find about 32,000 nearby potential customers, and optimize promotions through synthetic surveys, before posing the next challenge: scaling the technology from 40,000 employees to 5,000,000 partners.