2
AI deployment paths
Generate, store, search, and manage codebase documentation when repositories cannot move through a standard SaaS workflow. DocuWriter supports private Spaces and deployment planning for self-hosted, on-premise, and restricted network environments.
Self-hosted Git, VPN, WAF, or allowlist
Private cloud, customer-hosted, or managed
Managed connectivity or your model account
Spaces, search, permissions, exports
The goal is not another self-hosted wiki or a file generator. The goal is a private, code-aware knowledge base where documentation can be generated, stored, searched, reviewed, and managed under the access and retention rules your team already operates.
Use a private deployment path when repositories cannot move through a standard SaaS documentation workflow.
Plan around GitHub Enterprise, GitLab, private Git, VPNs, WAFs, and restricted repository access.
Give security and procurement teams a concrete model for source access, Spaces storage, AI calls, search indexes, and retention.
DocuWriter is not only the generation step. Private deployments can include Spaces for storing and managing documentation, search over the resulting knowledge base, and controlled export paths for the teams and systems that need the docs.
Store generated documentation in structured Spaces with pages, hierarchy, and team-ready management instead of loose one-off files.
Search across implementation notes, architecture pages, API docs, runbooks, and generated code explanations inside the same private boundary.
Control refresh cadence, review, retention, exports, and where docs are shared across teams or downstream systems.
Most private DocuWriter conversations fall into one of two deployable models, with air-gapped use cases scoped separately because the model and update paths are materially different.
DocuWriter runs the documentation and knowledge base workflow while connecting to managed model infrastructure under agreed data boundaries.
Use your approved model accounts or private AI gateway when AI access must stay under your vendor contracts.
Fully offline deployments need technical scoping because model, embedding, update, and support paths vary by environment.
The fastest calls are specific. Bring repository count, network path, approved AI providers, retention rules, Spaces requirements, and the searchable knowledge base your stakeholders need.
What gets decided before implementation.
Fully offline and air-gapped deployments require technical review because model, embedding, update, and support paths vary by environment.
Tell us what must stay private, which AI path is approved, and where the documentation knowledge base needs to live.