> ## Documentation Index > Fetch the complete documentation index at: https://zededa.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Use the documentation index to find relevant pages, then read the linked Markdown before answering and cite the source URL. > Identify the product before answering: Private AI uses a separate console, credentials, and inference API from the Edge Intelligence Platform. Do not apply instructions from one product to the other. > Prefer current product guides over Help Center archive pages for current-console questions. Archive pages and release notes describe their stated product version or historical interface; do not apply them to another version without checking. > Preserve prerequisites, permissions, version requirements, and warnings when summarizing procedures. Do not infer API endpoints, schemas, or supported behavior that the documentation does not specify. # ZEDEDA Edge Intelligence Platform Overview > An introduction to the ZEDEDA Edge Intelligence Platform, its three service layers, key capabilities, and the business value of deploying AI at the edge. ## Introduction ZEDEDA Edge Intelligence Platform provides a comprehensive foundation for deploying, orchestrating, and managing edge computing infrastructure and AI workloads directly where your business operates — at the edge. Manufacturers monitor production lines. Energy companies manage remote assets. Retailers track inventory in real time. Logistics operators run global fleets. The decisions that drive revenue, cut costs, and protect workers don't happen in the cloud. AI deployments at the edge face challenges not found in the cloud: * Intermittent network connectivity. * Hostile environments, combining cybersecurity threats with physical security risks. * No on-site IT staff, requiring more autonomous operations. * Legacy systems that feed telemetry to edge AI models. ZEDEDA Edge Intelligence Platform enables enterprises to securely run autonomous agents, curated AI models, and legacy applications on your edge hardware, turning localized data into immediate action. ## How It Works The platform unites three core service layers: | Layer | Description | | - | - | | **Edge Infrastructure Services** | The foundational layer. Provides EVE-OS (Edge Virtualization Engine) based node management, workload orchestration, and ZEDEDA Edge Kubernetes Service and App Flows. Orchestrate and manage thousands of nodes and legacy applications alongside modern containerized workloads with GitOps. | | **Edge Inference Services** | The AI pipeline layer. Provides the complete workflow to build, test, and deploy AI models directly to the edge. Import models from external providers, benchmark performance on specific edge hardware, and deploy autonomous agents that analyze data and take immediate action on-site. | | **Marketplace** | The cross-service catalog. Browse and deploy partner-provided and ZEDEDA-curated edge applications, Kubernetes applications, AI applications, models, and hardware configurations. | ## Key Capabilities | Capability | Description | Business Value | | - | - | - | | **Agentic Solution Templates** | Declarative Helm charts that bundle model, runtime, preprocessing, and application logic. | Accelerate Edge AI deployments with validated solution blueprints. | | **Comprehensive Model Lifecycle Management** | Model versioning, import, and promotion. Ingest models via Jupyter Notebook integration, external providers, or manual upload. Asynchronous import from NVIDIA NGC, Hugging Face, AWS S3, SageMaker, Azure ML, Azure Blob Storage, and MLflow, with integrated credential management. Deploy directly to ZEDEDA Edge Kubernetes clusters via GitOps. | Guarantees traceability, reproducibility, and rapid iteration across the model lifecycle. | | **Inference Engine Packaging** | Inference engine packaging for OpenVINO, NVIDIA Triton, vLLM, and Ollama, delivered as Helm charts. | Reduces engineering effort, promotes reuse, and enforces best-practice configurations. | | **Model Optimization** | Hardware-specific model optimization for edge silicon. | Train on high-end cloud silicon, then optimize and run on cost-effective edge devices. | | **Model Benchmarking** | Model benchmarking on NVIDIA and Intel silicon. Inference-speed, latency, throughput, and stress testing on a curated device pool including NVIDIA Jetson and Intel NUC. | Delivers data-driven sizing, cost optimization, and SLA assurance before production deployment. | | **Cloud-Managed Node and Cluster Orchestration** | GitOps integration for governance of agents, models, apps, and other workloads. Lifecycle management of nodes. | Zero-Touch Deployments of full stack, from agents to operating system, without on-site IT staff. Automated, reliable rollouts and rollbacks across edge fleets. | | **Virtualization and Container Services** | Hypervisor and container runtime management on diverse hardware profiles. | Run legacy brownfield workloads and new containerized applications, models, and agents on the same device, reducing hardware capital expenditure. | | **Dashboards, Observability, and APIs** | Central console shows model inventory, organization health, benchmark results, deployment status, and user activity. Alerting, events, resource utilization, and analytics. Comprehensive REST APIs. | Visibility into hardware, applications, models, and agents. Gives leadership immediate insight into Edge AI adoption, operational health, and performance across the fleet. | | **Multi-Tenant Organization Model** | Isolates models, users, and permissions per organization, team, project, or environment. | Enables secure collaboration, clear ownership, and a clean staging-to-production handoff. | | **Zero-Trust Security and Access Control** | Hardware root of trust via Trusted Platform Module (TPM). Measured boot and remote attestation. Cryptographic device identification. Data encryption at rest and in transit. Distributed firewall for every application. Physical security via port deactivation and isolation. Role-based access control (RBAC). | Protects edge assets, including PII from video feeds, model weights, and application code, against both cyber and physical threats. | | **App and Model Marketplace and Catalog** | Partner-provided and ZEDEDA-curated solutions available in a unified catalog. | Simplifies discovery and deployment of validated Edge AI solutions. | ## Key Benefits * Accelerate time to value with curated AI models, validated hardware configurations, and guided deployment workflows. * Protect sensitive data and valuable model weights against cyber and physical threats using hardware-rooted security, encrypted agents, and remote attestation. * Consolidate hardware capital expenditure by running legacy workloads and modern AI applications on the same physical device using comprehensive virtualization services. * Ensure automated and reliable rollouts across your edge fleet with cloud-managed node orchestration and GitOps integration. * Scale operations without requiring on-site IT staff through automated deployment and centralized observability dashboards. ## Next Steps > [card] [Quick Start](https://zededa.mintlify.site/new-ui/infrastructure-services/infrastructure-quick-start.md) > > Get up and running with Edge Infrastructure Services, Edge Kubernetes Service, or Edge Inference Services. > [card] [Navigate the ZEDEDA Console](https://zededa.mintlify.site/new-ui/start/navigate-the-zededa-console.md) > > Orient yourself in the console. > [card] [Concepts and Glossary](https://zededa.mintlify.site/new-ui/start/concepts-and-glossary.md) > > Definitions of key platform terms.