Technical Architecture
Digital Twin Pro is built as an appliance platform. The NVIDIA Jetson hardware provides the edge-AI foundation, while Digital Twin Pro adds the preinstalled assistant, pairing flow, dashboard, care plans, update path, support process, and client-specific setup. The hardware is the foundation; the managed private AI experience is the product.
The stack
- Hardware foundation — NVIDIA Jetson Orin Nano: up to 67 INT8 TOPS, NVIDIA Ampere GPU with 1,024 CUDA cores and 32 Tensor Cores, 8GB LPDDR5 at 102 GB/s, operating at 7W–25W. Housed in a ROBOCO precision aluminum cooling case with a preloaded 1TB NVMe drive.
- System image — built on NVIDIA JetPack 7.2, flashed, configured, and tested before shipment.
- Hermes — the preinstalled personal AI assistant, tuned per client before the appliance ships.
- OpenClaw (optional) — an agent framework available for advanced users on request.
- Agent runtime direction — our roadmap follows NVIDIA's NemoClaw reference stack for running always-on agents inside OpenShell sandboxes: guided onboarding, routed inference, network policy, and lifecycle management for safer managed agent workflows.
- Local model runtime — capable local models for private everyday work, with optional bring-your-own-key connections to Claude, GPT, Gemini, and other providers when you choose.
- Pairing & dashboard — QR-based phone pairing; a browser dashboard for connections, settings, and care.
- Care layer — optional managed updates, health monitoring, encrypted backups, and human support.
- Private large-job processing — for approved archive jobs too large for the appliance, private processing returns the finished results to your device.
Hermes vs OpenClaw — and what NemoClaw gives both
Hermes is the assistant every Digital Twin Pro ships with: configured for you before shipment, tuned to grow with how you work, and supported by our care plans. OpenClaw is an optional agent framework we can configure for advanced users who want to build their own agent workflows.
Both run within NVIDIA NemoClaw — NVIDIA's open-source stack for always-on edge agents — which provides the same foundation to either agent: OpenShell sandboxing for isolation, guided onboarding, routed inference across local models, network policy, and lifecycle management. In practice: Hermes if you want a finished assistant; OpenClaw if you want a framework; NemoClaw keeps either one safe, managed, and current.
What this means in practice
You are not buying a developer kit. A Jetson developer kit is a powerful platform for engineers. Digital Twin Pro is a finished private AI appliance: assembled, configured, preloaded, paired from your phone, supported by optional care plans, and designed for non-technical owners — with the technical depth available to those who want it.
NVIDIA, Jetson, and JetPack are trademarks of NVIDIA Corporation. Digital Twin Pro is an independent product company and is not affiliated with, endorsed by, or sponsored by NVIDIA unless otherwise stated.