> ## Documentation Index > Fetch the complete documentation index at: https://docs.revoengine.com/llms.txt > Use this file to discover all available pages before exploring further. # Choose your approach > Understand when RevoEngine fits, select capabilities, and separate platform responsibilities from application decisions. RevoEngine is an enterprise Backend as a Service and governed-execution platform for building, integrating, operating, and improving production applications. It combines application logic, data, Storage, HTTP endpoints, automation, identity, observability, and AI in one instance-scoped application model. Low-code is one authoring option, alongside custom Node.js and external developer tools. Use RevoEngine when a team needs to deliver and operate business applications or integrations with shared identity, access control, versioned definitions, execution history, and operational ownership. Its value is the connection between a business capability, its implementation, the authority under which it runs, and evidence of the outcome. ## What the platform owns and what you own * The control plane manages definitions, versions, access, and operational state. The execution plane runs the selected workload through its own synchronous, transient, or durable execution contract. * The platform supplies managed capabilities and their controls. The application team owns business rules, data quality, idempotency, external-system contracts, rollout decisions, and business success criteria. * Platform UI, Platform API, SDK, CLI, and AI are access paths to the same application model. An interface choice does not create a separate source of truth or bypass permissions. * Documentation explains supported product concepts and workflows. Verify the target instance's current contract, enabled capabilities, permissions, and version before execution; latest documentation is not evidence of deployment. * A definition, accepted request, completed transfer, or dispatched event is not proof of business completion. Follow the relevant durable result, execution status, or finalized artifact. * RevoEngine does not automatically replace an existing system of record or design the application's business rules. Integrate with the existing owner when it already provides the required capability. ## Choose capabilities for the outcome ### Publish a business operation An Endpoint backed by a validated Component for bounded synchronous work. For long work, accept a command and follow a durable Job result. **Check before implementing:** Input validation, identity, expected duration, duplicate requests, and the caller's completion contract. [Read the guide](/operate/endpoints) ### Import or export files Storage for durable inputs and outputs, with an upload session for transfer and bounded structured reads for CSV, TSV, or XLSX. Use a Job when processing must outlive the request. **Check before implementing:** Finalization, schema and parsing policy, access, retention, and idempotency by source identity and version. [Read the guide](/operate/storage) ### Connect an existing business system Keep the existing system of record where appropriate. Use Components, HTTP integration, events, or webhooks to implement the integration boundary. **Check before implementing:** Ownership of each field, credentials, retry safety, reconciliation, and external rate or execution limits. [Read the guide](/low-code/http) ### Choose an authoring interface Use Platform UI or Web IDE for in-platform work, CLI and VS Code for local development, and SDK or Platform API for programmatic integration. **Check before implementing:** Whether the task manages definitions, invokes an Endpoint, or executes runtime code; these are different contracts. [Read the guide](/developers/overview) ### Investigate a production failure Start from the exact resource or execution and its operational evidence; relate the failure to configuration and version before proposing a change. **Check before implementing:** The original error, whether side effects happened, the affected scope, and proof of recovery. [Read the guide](/operate/observability) ### Choose deterministic automation or AI Use deterministic Components and Jobs for known business rules. Use an Interactive or Autonomous Agent when interpretation, investigation, or adaptive capability use is required. **Check before implementing:** Acceptance criteria, authorized capabilities, execution identity, approvals, cost and execution bounds, and durable evidence of completion. [Read the guide](/ai/overview) ## How the capabilities fit together | Capability | Responsibility | Guide | | ----------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------- | | Application logic | Implement reusable business behavior in Components and libraries; validate it in Sandbox before selecting a production version. | [Application logic](/build/components) | | Endpoints | Expose a bounded HTTP contract over application logic when the caller needs a synchronous result. | [Endpoints](/operate/endpoints) | | Data | Model application-owned structured data, query it through Tables and Views, and track changes with the appropriate access and audit policy. | [Data](/operate/databases) | | Storage | Keep files, generated outputs, and evidence addressable by stable entry identity. Distinguish an entry from bytes, an upload session, and a derived index; consumers require a finalized file. | [Storage](/operate/storage) | | Automation | Run durable background work. Job templates define work and Jobs record executions; schedules and events are triggers that can target a Job template or an Agent where enabled. | [Automation](/operate/automation) | | Identity and governance | Assign human and service identities, roles, Groups, ACLs, and Secrets so each action has an explicit authorized scope. | [Identity and governance](/operate/identity-and-access) | | Operations | Connect changes to runtime outcomes using Activity, versions, logs, traces, audit, and execution history. | [Operations](/operate/observability) | | AI | Use Interactive Agents for supervised work and Autonomous Agents for durable responsibilities. Both run through the governed Agent Harness with explicit identity, capabilities, approvals, memory, evidence, and ownership. | [AI](/ai/overview) | ## Move from a design to verified work State the business outcome, existing system of record, identity, expected volume, completion boundary, and failure policy. Choose the smallest combination of platform capabilities that meets those requirements. Inspect the exact contract and validate both success and failure paths before production use. Continue with [Secure order processing](/platform/first-production-workflow) for a complete example, or [Developer overview](/developers/overview) to choose an interface.