NeduraTalk to us ↗

DEPLOYMENT & GETTING STARTED

Your infrastructure.
A path to first use.

Nedura is installed and upgraded by your administrator using nedura-deploy. Plan the environment, verify a provider connection, and build from one working conversation.

01 / DEPLOYMENT PATHS

Choose the environment you operate.

LinuxWindows ServerContainers / HelmTerraform

The installation documentation identifies administrator-managed deployment paths. To evaluate Nedura, request evaluation access and we'll send download access and setup steps.

Work with your infrastructure team to review the deployment approach, service configuration, model access, credentials, and required network connections.

Installation documentation ↗

02 / SIGNED RELEASES

Signed releases

Each Nedura release is signed and ships with a software bill of materials (SBOM). nedura-deploy checks the installed packages against the SBOM; on Nedura Gov a mismatch stops the install.

Upgrades keep the previous version on disk for rollback, and configuration changes are applied with nedura-deploy reconfigure.

Approved clients receive releases from ALINEDS.

03 / FIRST WORKING SESSION

Keep the first milestone simple.

  1. Complete administrator installation

    Use nedura-deploy for the selected environment and consult current installation guidance.

  2. Configure a model provider

    Use the setup and model workflow. For custom endpoints, confirm the endpoint, model name, and context settings match the server.

  3. Verify a normal conversation

    Complete a basic request before adding gateways, fallback chains, voice, or schedules.

  4. Add tools for one real task

    Enable a focused toolset, choose relevant skills, and validate the result with a task your team can review.

  5. Connect the delivery surface

    Configure the messaging gateway, API client, or editor after the underlying conversation and tools work.

Follow the quickstart ↗

04 / CONFIGURATION

Prepare for how the agent will work.

Profiles

Use separate profiles for independent roles. Each has its own configuration and persistent state.

Profile guide ↗

Models and endpoints

Confirm credentials and model availability. Add optional fallback providers after the primary connection works.

Fallback guide ↗

Tools and approvals

Review enabled capabilities, tool backends, MCP access, and approval checks for interactive and unattended work.

Security guide ↗

05 / BEFORE EXPANDING

Validate the complete workflow.

Check tool execution, session continuity, result delivery, and any approval steps. Test scheduled tasks with their configured output destination.

Identify the failing layer when troubleshooting: model authentication, gateway connections, and tool backend errors need different checks.