Enterprise overview
Afunana brings system knowledge into an environment your teams can manage. Deployment, source access, AI use and oversight are choices to agree as part of a focused evaluation.
01 / Deployment
The application runs on your infrastructure using Docker or Podman. Use the bundled database and web proxy, or integrate with a customer-managed SQL Server and reverse proxy. Offline installation options support environments with restricted connectivity.
Source, generated documentation and the search index are held in your deployment. Search embeddings are computed locally. If you configure a cloud AI provider, relevant source and context can be sent to that provider for analysis. Local-model configurations support AI processing within your own network.
For the POC: agree the source scope, model provider, outbound connections, retention settings and operational owner before starting the analysis.
02 / Identity and access
Afunana supports enterprise sign-in through SAML and OIDC, alongside local accounts. Collection-level access helps separate bodies of analyzed source for different teams. Administrative functions are distinct from day-to-day knowledge exploration.
Define the roles and collections needed for your deployment, then verify the permitted actions with representative users. Include browser access, developer tools and integrations in that review so the evaluated boundaries match how your teams will work.
A POC can focus on extraction and analysis without enabling write-back. Optional change execution uses a separate approval workflow. Developers remain responsible for compiling, testing and accepting any resulting code change.
03 / AI governance
Configure AI connections and assign models to tasks such as documentation and chat. Supported options include cloud providers, local models and organization-managed gateways. Provider choice can change without changing the analyzed application's source.
Model-call records and usage metering provide visibility into the models used and their calculated cost. Approved-model settings, review workflows and budget controls can be configured to support the organization's policies.
Afunana's governance features concern its own use of AI. Your organization defines approvals, responsible owners, data-handling requirements and review practices. Product records can support that work; they are not a substitute for an organization-wide AI governance program.
04 / Evidence and review
Documentation and answers include source references that specialists can follow. Dependency maps and structural checks provide a different kind of evidence from AI-written interpretation. On IBM i, dependency verification can compare documented calls with the platform's cross-reference.
Audit logging and configurable forwarding support the review of security-relevant activity. Model usage and governance records help reviewers understand how the configured system is being used. Agree event coverage, retention and reporting requirements as part of the evaluation.
Standards and certification: the product includes control mappings and assessment features to support review. These are not a claim of SOC 2 or ISO 27001 certification. Discuss current assurance materials and your specific requirements with the team.
05 / Working with the knowledge
The web application supports top-down exploration, business documentation and questions. The VS Code extension brings documentation and source context into the developer's editor. Authenticated interfaces let supported AI assistants and other integrations query the analyzed knowledge.
Specifications, manuals and team notes can join the searchable knowledge base. Exportable documentation helps teams share findings with stakeholders who do not work in the application every day.
A practical next step
Bring your application questions and operating requirements. Together, we can define a POC that gives both the business and technical teams useful evidence.
