Customer-site network and access guide
A plain-language map of where the appliance sits, which network paths it uses, and how office and remote employees connect.
Open resourceResources
Make the deployment understandable before it becomes technical.
Use these guides and working sessions to align business, IT, security, and facilities stakeholders around the same system boundary.
Buyer and operator resources
Answers you can put in front of the whole room.
Available resources open directly. On-request worksheets are delivered through a short discovery exchange so they can be matched to the engagement.
Private AI readiness checklist
The users, workloads, facilities, identity, security, and support decisions to settle before hardware is ordered.
Request resourceDeployment and acceptance overview
What happens from discovery through staging, installation, verification, operator training, and handoff.
Open resourceWorkload discovery worksheet
A structured way to identify the first workflows worth testing and the evidence required to approve them.
Request resourceCommon questions
The practical answers.
Every final design is customer-specific. These answers establish the default operating pattern.
Where does the hardware sit?
At the customer site, normally in a locked server room, network closet, or other controlled IT space with appropriate power, cooling, and physical access. Tower and rack options depend on the selected tier and facility.
Is SkilakSpool on the local network or the public internet?
It is connected to customer-approved local network segments and presented through private HTTPS. The user interface is not intended to be published directly to the internet, and backend model services remain restricted.
How do employees working from home connect?
Through the customer’s managed remote-access path—typically a managed device, MFA, and VPN or zero-trust network access. Remote access is designed with the customer’s IT and security teams, not added as a separate unmanaged tunnel.
Can it run without internet access?
Yes, the core AI workload can be designed for offline operation. Cloud identity, software updates, monitoring, or remote support may require narrowly scoped outbound connectivity; those paths are optional, documented, and customer approved.
Do you stage the system before delivery?
Yes. Skilak configures and exercises the appliance on a controlled staging network using synthetic or explicitly approved data, then transports it to the customer site for final network, identity, and acceptance work.
Who can access the appliance?
Only users and administrators authorized by the customer. The design separates normal use from administration and can integrate enterprise identity, roles, and audit events where required.
Does SkilakSpool make us compliant?
No product alone creates compliance. SkilakSpool is designed to operate within the controls, policies, evidence, and responsibilities of the customer’s broader security program.
Still mapping it?
Put the open questions on one page.
A discovery briefing is built to clarify the physical site, users, data, access, workload, and approvals—not force a premature hardware decision.