AI for operations: PwC France pilots a ‘world model’

PwC France and Maghreb said on April 15 that it will use ManakoAI’s “Business Operations World Model” to help clients model and automate work on shop floors, in warehouses and across logistics networks. (manako.ai) The pitch is simple: use existing camera feeds as live inputs, then let software detect what is happening and trigger actions such as dispatches, audit reports or workflow automations. Manako said the system is aimed at retail, logistics, manufacturing, energy and infrastructure clients. (manako.ai) A “world model” in this context is software that builds a running picture of a physical operation, like a digital control room that updates from video and sensor data. Manako said PwC France will rely on that layer rather than ask clients to build their own specialist computer-vision stacks in house. (manako.ai) The technical plumbing runs through Score, which Manako described as a decentralized computer-vision network, and Score is tied to Bittensor’s subnet structure. Bittensor says its network had 128 active subnets on April 17 and pays participants in TAO tokens for AI work such as model training, inference and evaluation. (manako.ai) (bittensor.ai) PwC is not presenting this as a standalone lab experiment. Manako said the alliance is “anchored in PwC France” but intended to operate at global scale through PwC’s wider network of member firms. (manako.ai) That fits PwC France and Maghreb’s recent push to move from internal generative artificial intelligence adoption to client-facing deployments. In June 2025, PwC France said it had deployed more than 6,000 GenAI licenses internally and reached an 80% adoption rate. (pwc.fr) The sectors named in the alliance are also the ones under pressure to react faster to disruptions in transport and supply chains. PwC’s global transportation and logistics practice says companies face higher costs, more complex trade routes and a need for “sharper capabilities” as global trade shifts. (pwc.com) The open question is whether companies will trust a blockchain-linked AI stack for operational decisions that affect safety, shrinkage and downtime. Manako’s announcement describes the system as cost-efficient and deployable on existing infrastructure, but it did not disclose client names, contract values or measured results from live deployments. (manako.ai) What PwC France has put in motion is less a chatbot for office work than software for watching the physical world and acting on it. The next test is whether that “system of action” can move from pilot language to repeatable results inside retail, factory and logistics operations. (manako.ai)