Day 1
Start validating AI on your internal documents soon after install
Local Stack
Run the LLM, embeddings, and vector store on your internal network
Incremental
Re-index only changed files to stay current
AIAppliance
OnNetworkAI
OnPremise
DocumentRAG
SourceBackedAnswers
LocalLLM
IncrementalIndexing
Why Appliance
Start AI adoption as a product, not a project.
Wissly-in-a-Box isn't just a server — it's a complete RAG stack that runs inside your corporate network. It connects documents on local PCs, NAS, and internal servers, turning them into working knowledge.
Key Points
What makes Wissly-in-a-Box different
From fast adoption to network security, local data connectivity, a complete RAG stack, incremental indexing, and source verification — see the six points that make Wissly-in-a-Box different where the work actually happens.
Security
Sensitive documents stay on your network; AI stays close to where the work happens.
Wissly-in-a-Box supports the use of sensitive documents and AI-powered work through an operating model centered on your internal network.
Enterprise Network Boundary
On-premise
Local PC / NAS / Internal server
Contracts, technical docs, research, and customer-support documents that are hard to upload to the cloud
Users
Business users, administrators, and IT operators
Wissly-in-a-Box
Local LLM, embeddings, vector store, document indexing, RAG Q&A
Incremental indexing pipeline
Converts, parses, and indexes only changed files to stay current
Admin tools / Internal integration
Permissions, logs, system status, API, and business portal integration
Network-centric operation
Built around a setup where the local LLM, embeddings, and vector store all run inside your network.
Permissions and logs
The scope of user access, admin operations, and audit logs is finalized to match the actual supported spec.
Network separation & air-gap review
Air-gapped networks, offline setups, and update policies are confirmed per deployment environment.
Direct local data connection
Connects internal knowledge that cloud-only tools can't reach, such as local PCs, NAS, and internal servers.
Hardware Information
Hardware specification table
Organized around the finalized hardware configuration so buyers can compare at a glance. See the appliance's performance and operating scope in one table.
CPU
Xeon 24C / 48T
RAM
512 GB
GPU
RTX PRO 6000 × 2
Storage
4.5 TB
On-premise AI appliance
Corporate network / On-premise environment
Intel(R) Xeon(R) Platinum 8559C (24C / 48T)
NVIDIA RTX PRO 6000 Blackwell Server Edition × 2
512 GB RAM
EBS 1TB · NVMe Instance Store 3.5TB
eth0 10GbE
Document search, RAG Q&A, source verification, admin tools, incremental indexing
Local LLM, embedding model, vector store
Local PC, NAS, internal server, on-premise folder
Incremental indexing focused on changed files
Deployment
Deployment process
After reviewing your network environment and document storage structure, we proceed step by step through hardware configuration and initial document connection.
