Huawei Unveils New AI Data Platform to Accelerate Business AI Deployment

During the Huawei AI DC Innovation Forum at MWC Barcelona 2026, the company revealed its latest AI Data Platform, tailored to tackle major challenges hindering the scalability of AI solutions in enterprises.

Key Challenges Addressed

In the current landscape, AI agents are at the forefront of technological change. However, businesses still struggle with deploying these agents effectively, primarily due to issues like delayed data assimilation, low accuracy in search, and inefficient reasoning in scenarios involving long and multi-step interactions. Additionally, there is often a lack of task memory and experience accumulation, which keeps many AI agents in the demonstration stage, far from practical real-world applications.

Directly addressing these challenges, Xie Liming, the President of the Flash Storage Domain at Huawei, introduced the AI Data Platform. This platform integrates a knowledge base, a KV cache, and a memory bank, orchestrated through UCM, enabling companies to transition their AI agents from mere demonstrations into fully functional production tools.

Real-Time Knowledge Generation and Retrieval

The platform utilizes advanced technologies to generate and retrieve knowledge using precise, multi-modal solutions for agents. It employs knowledge bases to constantly detect changes in source data, seamlessly transforming raw data into actionable knowledge in near real time. Multi-modal data is transformed into high-accuracy knowledge through lossless multi-modal parsing and token-level encoding, achieving search accuracy rates exceeding 95%.

Accelerated Inference with KV Cache

The KV cache accelerates inference by leveraging data from historical memory, significantly reducing redundant computations during inference tasks. This optimization translates into lower latency, increased throughput, and improved user experience, especially in scenarios involving lengthy sequences and complex agent tasks.

Memory Extraction and Task Recall

Further enhancing performance, the platform features memory banks that collect both active and experiential memory during interactions with AI agents. It allows for memory history playback and enables collaboration between multiple agents in learning processes, continually improving the accuracy and efficiency of inferences. This means that the models become increasingly intelligent the more they are used.

Future Directions

Looking ahead, Huawei plans to strengthen its investments in data infrastructure for AI. The company aims to support industry transformations through continuous innovations and to collaborate with global clients and partners to accelerate AI deployment across various sectors. With these advancements, Huawei seeks to fully unlock the potential of data and drive efficiency in AI applications.

The AI Data Platform presents a promising step toward overcoming existing barriers in AI implementation, positioning Huawei as a pivotal player in shaping the future of intelligent enterprise solutions.