| 💬 Chat Q & ARetrieval-augmented chat over your own knowledge base, deployable end to end on a single Xeon node with one installer. | 📄 Document SummarizationSummarizes large documents and datasets, returning concise and relevant insights quickly. | 🎙️ Audio Q & AVoice in, voice out. Speech-to-text and text-to-speech wrapped around the RAG pipeline for dynamic spoken conversations. |
| 🌐 Multilingual SupportSwap in multilingual embedding, reranking and LLM models to run end-to-end RAG in 100+ languages. |
| ⚡ Intel® Inference ToolkitKubernetes-orchestrated LLM serving on Intel hardware. Automates model deployment, compute provisioning and tuning for faster inference with less manual effort. | 🤖 Intel® Agent ToolkitProduction agentic AI stack on Xeon: API gateway, inference routing, agent memory, sandboxed code execution, MCP tool integration and full observability, pre-integrated. | 🎬 Video Search and SummarizationSearch through video datasets and generate concise summaries. Deploys on Kubernetes with vLLM as the inference backend. |
| 🎩 Red Hat® OpenShift® AI on XeonInference and agentic AI on Red Hat OpenShift AI, with vLLM CPU tuning, KV-cache sizing and GuideLLM benchmarking guidance. | 🗄️ NetApp® AIPod™ Mini with IntelAir-gapped RAG inferencing over an internal knowledge repository, serving concurrent users without GPUs. | ☁️ Nutanix® Enterprise AI with IntelChat Q&A at scale using Nutanix Enterprise AI as the inference engine, on Nutanix Kubernetes Platform or any Kubernetes cluster. |
| 🖥️ VMware® vSphere with IntelProvisions a production-ready Kubernetes cluster on vSphere / ESXi and runs Chat Q&A on Xeon with Intel® AMX acceleration. | 🏭 Dell® On-Prem Automated DeploymentBare-metal automation via iDRAC Redfish and Terraform: custom ISO boot, then an automated post-OS script that launches inference services. | 🔌 Dell® Airgapped DeploymentStages Docker images, Helm charts, Python packages, binaries and model weights in JFrog Artifactory, then installs on a system with no internet access. |
| 🎮 Dell® with Xeon Host CPU and NVIDIA GPUPowerEdge deployment with Xeon as host CPU: installs the NVIDIA GPU operator and serves a model with vLLM on the GPU. |
| 🔐 Confidential ContainersConfidential AI powered by Intel® TDX. Protects models and data from compromised software and administrators, even on untrusted infrastructure. | 📈 Extension for scikit-learnFree AI accelerator delivering up to 100x faster performance for existing scikit-learn code, with no code rewrite required. |