Guides

Getting Started

apriti connects to your Salesforce org via JWT authentication, extracts metadata, computes dependency graphs, and generates documentation — all automatically.

1. Connect Your Org

Configure JWT Bearer authentication with your Salesforce org. Read-only access with a Connected App — no user passwords needed.

2. Run a Scan

Trigger a scan via the API or the back office. apriti fetches metadata, builds dependency graphs, analyzes automation paths, and generates documentation.

Architecture

apriti follows a WorkItem-driven architecture. Every scan is broken into discrete, idempotent work items that execute in sequence:

INIT_SCAN
  → FETCH_METADATA
    → BUILD_DEPENDENCIES
      → ANALYZE_METADATA
        → BUILD_DOCUMENTATION
          → FINALIZE_SCAN

Each work item is idempotent, has retry semantics, and can be resumed safely after failures. The system is multi-tenant with strict data isolation per tenant.

Output Formats

Markdown

Structured markdown files per object, process, and business rule. Ideal for Git-based documentation.

HTML

Standalone HTML pages with navigation, cross-references, and search-friendly structure.

Frequently Asked Questions

Can't find what you're looking for? Reach out to contact@aprity.ai.

How does apriti connect to my Salesforce org?

apriti uses Salesforce JWT Bearer authentication for secure, read-only access. No passwords are stored — only a private key for token-based authentication. Zero footprint on your org.

Is the analysis AI-based or deterministic?

The core analysis (metadata extraction, dependency graph computation, impact analysis) is fully deterministic — no AI involved. AI is used only for the documentation generation step: explaining business rules and processes in plain language.

What output formats are supported?

apriti generates documentation in Markdown and HTML. These can be published to internal wikis, exported as PDF, or integrated into your existing documentation systems.

What languages are supported for documentation?

Documentation can be generated in 6 languages: English, French, Spanish, German, Italian, and Portuguese.

How long does a scan take?

Scan duration depends on the size of your org. A typical org with a few hundred custom objects completes in minutes. Large enterprise orgs with thousands of metadata items may take longer. Scans are incremental — subsequent runs are faster.

Where is my data hosted?

apriti runs on Microsoft Azure with EU data residency (France Central). Your Salesforce metadata is processed in Azure and stored in Azure Cosmos DB with strict tenant isolation.