Huawei Launches AI Data Platform to Accelerate Enterprise AI Adoption

Barcelona, Spain – at Huawei AI DC Innovation Forum held during MWC Barcelona 2026, Huawei unveiled the AI Data Platform, a platform designed to tackle a range of challenges in deploying AI agents and to strengthen the data foundations for digital transformation and intelligence within companies. Today, AI agents sit at the centre of business transformation. Despite having large volumes of data, many companies still struggle to deploy AI agents at scale. Key challenges include delays in knowledge acquisition, low data search accuracy, inefficient inference processes in lengthy, multi-round interactions, and lack of task memory and accumulated experience. These constraints leave most AI agents largely at demonstration stage and not ready for full enterprise deployment. In response to these challenges, Xie Liming, President, Flash Storage Domain, Huawei Data Storage Product Line, introduced the AI Data Platform. The platform integrates a knowledge base, KV cache, and memory bank, and is managed via UCM. With this approach, a company’s AI agents can evolve from demonstrations to tools truly used in operational activities. - Building and querying a knowledge base with high accuracy in real-time for AI agents This technology uses a knowledge base to continuously detect changes in source data and convert raw data into near-instant knowledge. Multimodal data can also be processed through parsing without information loss and token-level programming, resulting in a knowledge base with a search accuracy of over 95%. - KV cache to accelerate inference, leveraging historical memory data to improve AI agent inference efficiency By leveraging historical memory data, a smartly managed KV cache can reduce repeated computations during inference. This lowers latency, increases throughput, and provides better performance support for complex AI agent inference. - Memory extraction and retrieval for AI agents with personalised, continuous memory summaries Through the memory bank, this technology collects working memory and experiential memory during AI agent interactions. The system supports memory retrieval and collaborative learning among agents so that accuracy and inference efficiency can continuously improve with use. Going forward, Huawei will continue increasing investment in AI data infrastructure, driving industrial transformation through sustained innovation, and collaborating with customers and global partners to broaden AI deployment across sectors and maximise the potential of data.