Intel has unveiled collaborations with the International Olympic Committee (IOC) and introduced a generative AI (GenAI) retrieval-augmented generation (RAG) solution, emphasizing open and accessible AI systems.
Athlete365: AI-Powered Support for Olympic Athletes
In partnership with the IOC, Intel developed Athlete365, a chatbot powered by Intel Gaudi AI accelerators and Xeon processors. This chatbot assists the 11,000 athletes at the Olympic Games, providing on-demand information and helping them navigate logistics and regulations, allowing athletes to focus on their performance.
Tackling GenAI Deployment Challenges
Intel’s RAG solution addresses the challenges of deploying GenAI, such as cost, scalability, accuracy, and privacy. By securely leveraging proprietary data, Intel enhances the reliability of AI outputs, benefiting various industries. Intel’s collaborative approach, using open standards and a robust software ecosystem, enables developers to build customized GenAI RAG solutions tailored to enterprise needs.
GenAI RAG Solution Architecture
Built on the Open Platform for Enterprise AI (OPEA) foundation, Intel’s GenAI solution is flexible and customizable, integrating components from multiple industry partners. It scales with orchestration frameworks like Kubernetes and Red Hat OpenShift, offering standardized APIs with security and telemetry.
Open Software Stack and Future Collaborations
Intel’s collaboration with OPEA has produced an open software stack for RAG and LLM deployment, utilizing PyTorch, Hugging Face, LangChain, and Redis Vector database. This stack simplifies development on Intel AI systems.
Looking ahead, Intel is part of the newly formed Coalition for Secure AI (CoSAI) with Google, IBM, and others, focusing on trust and security in AI development. Intel will showcase its AI innovations and collaborations at the Intel Innovation event on September 24-25.
By fostering an open ecosystem, Intel is pushing the boundaries of AI while making advancements accessible to a broader range of developers and enterprises.
