How do you move from AI POCs to governed, profitable solutions in production? In this session recorded at the Tech for Industry Show, André Queiros (Manufacturing Lead France, OutSystems) and Floriane Mignon (Head of Product, Suez Digital Solutions) share hands-on feedback from the digital transformation of the waste and water industry.
The talk covers the four main barriers to scaling enterprise AI (rigid core systems, business semantics, governance, LLM vendor lock-in) and Suez's answer with the OutSystems low-code platform — development 3x faster than traditional coding, only 10% maintenance anomalies, and a jump from 2 to 20 applications in one year without expanding the development team. On the field: computer vision above 85% accuracy to monitor waste stream quality, AI-powered OCR reaching 98% reliability for customer invoicing, industrial IoT (fill-level sensors, container trackers) and 40% efficiency gains on selected traceability lines. The flagship demo: AI agents acting as guardrails around an internal LLM (GPT, Mistral) to simulate the closure of an industrial site while enforcing regulatory constraints — first queries delivered within 3 months.
A must-watch for CIOs, operations and business leaders looking to deploy sovereign, governed, value-driven AI — at the heart of Industry 4.0 and the circular economy.
