Documentation
Prava AI documentation
Where to start with Prava AI: setting up a workspace, deploying the Agent, using each module, and building on the API.
Get started
From sign-up to first detection
Six steps take a new workspace to live coverage. Most teams complete them in this order.
Step 1
Create a workspace
Sign up and choose a plan. The first user becomes the workspace owner. See pricing for plan limits.
Step 2
Choose your region
Select the region where your data will be stored and processed. It is set at onboarding and determines your regional endpoints and API base URL.
Step 3
Invite users and set up SSO
Invite your team, assign roles, and connect your identity provider through SAML or OIDC. Provision users automatically with SCIM.
Step 4
Deploy Agents
Install the Agent on endpoints and servers through your device management tool, a script or a package. Start with a pilot group.
Step 5
Connect sources
Add cloud accounts, email, identity, log sources and other systems through connectors, syslog/CEF, STIX/TAXII or the API.
Step 6
Enable modules
Turn on the modules you need (such as endpoint protection, dark web monitoring, VoIP security, satellite comms, vehicle systems, or phishing defence), set response policies, and decide which actions Prava AI may take automatically, with approval, or only recommend.
Guides
Guides by module
Each module has its own guide inside the platform. Here is what each one covers.
01
Endpoint protection
Agent telemetry, behavioural detection and containment on endpoints. Overview.
02
Application, server and cloud monitoring
Monitoring of applications, servers and cloud accounts for misconfiguration and suspicious activity.
03
IoT/OT monitoring
Visibility and anomaly detection for connected and operational technology devices.
04
VoIP and telecommunications
SIP and RTP signaling inspection, PBX vulnerability hardening, toll fraud prevention, eavesdropping defense, and carrier trunk security.
05
Satellite and RF communications
SATCOM terminal posture, VSAT uplink and downlink exposure, ground-station link defense, satellite telemetry stream auditing, and RF signal anomaly detection.
06
Connected vehicles and automotive systems
In-vehicle CAN bus and automotive Ethernet anomaly detection, telematics control unit (TCU) security, infotainment isolation, and fleet firmware integrity compliant with UN R155 and ISO/SAE 21434.
07
Penetration testing and attack surface
AI-guided testing, attack-path validation and continuous discovery of exposed assets. Overview.
08
Threat intelligence
Adversary tracking, campaign intelligence and curated threat feeds mapped to your external assets. Overview.
09
Dark web monitoring
Continuous surveillance of underground forums, illicit markets, ransomware leak sites and pastebins for compromised credentials and exposed enterprise data. Overview.
10
Phishing defence
Intent-based analysis of email for phishing, business email compromise and fraud. Overview.
11
Malware analysis
Static and dynamic analysis, reverse engineering, and YARA and Sigma rule output. Overview.
12
Forensics and incident response
Evidence collection, timeline reconstruction and supervised response to incidents.
13
Compliance and policy management
Mapping controls to frameworks, managing policies and tracking evidence.
14
Prava AI assistant
Natural-language questions, investigation summaries and reports across your tenant.
Agent
Agent deployment
The Agent is a lightweight sensor for endpoints and servers. It collects telemetry and carries out the response actions you allow.
Platforms
Supported operating systems
Current versions of Windows, macOS and major Linux distributions for desktops, servers, embedded systems, connected vehicle telematics units (TCUs), and edge gateways. Supported versions: [Supported OS versions and architectures].
Install
Installation methods
Deploy through your MDM or device management tool, run the install script, or install the signed native package directly. Each method uses a workspace enrolment token.
Network
Network requirements
The Agent needs outbound TLS (port 443) to your regional endpoints across terrestrial, cellular (4G/5G), and satellite links. No inbound ports are required. Proxy support: [Proxy configuration details].
API
API overview
Everything you can do in the console you can automate. The API is REST over HTTPS and uses JSON for requests and responses.
Requests authenticate with an API key or with OAuth 2.0 client credentials. Create keys in the console, scope them to the permissions they need, and rotate them regularly. Each region has its own base URL, so requests reach the data centre where your data lives.
# List recent high-severity alerts (illustrative)
curl -s "https://api.<region>.prava.dev/v1/alerts?severity=high&limit=50" \
-H "Authorization: Bearer $PRAVA_API_KEY" \
-H "Accept: application/json"
Note. This example, the endpoint names and the base URL pattern
https://api.<region>.prava.dev/v1 are illustrative. The exact endpoints, regions and
parameters are in the API reference available to customers inside the platform.
Pagination
Cursor-based pages
List endpoints return a page of results and a cursor for the next page. Set the page size with a limit parameter.
Errors
Standard status codes
Errors use standard HTTP status codes and a JSON body with a machine-readable code, a message and a request ID to quote to support.
Limits
Rate limits
Requests are rate-limited per key. Responses carry headers showing your remaining allowance, and a 429 status tells you to back off. Limits: [Rate limits by plan].
Events
Webhooks
Subscribe to events such as new alerts or completed actions. Deliveries are signed so you can verify they came from Prava AI, and failed deliveries are retried.
Versions
Versioning
The major version is in the path, such as /v1. Breaking changes come only with a new version, announced in advance in the changelog.
Regions
Regional base URLs
Use the base URL for your workspace's region. Regional hostnames: [List of regional API hostnames].
API usage by plan is described on API pricing.
Integrations
Connecting other tools
Prava AI connects to other systems through standard methods: the REST API, webhooks, syslog/CEF, STIX/TAXII, SAML/OIDC and SCIM. Dedicated connectors support cloud services, VoIP PBX and SIP trunks, SATCOM uplinks, and vehicular telematics systems. Availability of specific connectors varies by plan; confirm with us. The integrations page lists the methods and common targets.
Administration
Admin and identity
Roles
Role-based access
Assign roles such as owner, administrator, analyst and read-only. Roles control what users can see, change and approve, including autonomous actions.
SSO
Single sign-on
Connect your identity provider with SAML 2.0 or OpenID Connect, and enforce sign-in through it for all users.
SCIM
User provisioning
Create, update and remove users and group memberships automatically from your identity provider with SCIM.
Audit
Audit logs
Every sign-in, configuration change and autonomous action is recorded. View audit logs in the console or export them through the API.
Data
Data and privacy for developers
Data sent through the API or connectors is treated as customer data. It is stored and processed only in your region, and the same safeguards apply as for data collected by the Agent. Do not send secrets or credentials in free-text fields.
How our models use data is set out in the AI and Data Use Policy. Our processor terms are in the Data Processing Addendum.
Support
Changelog and help
Product and API changes are published in the changelog inside the platform. Deprecations are announced there before they take effect, and security-relevant changes also appear on our security page.
For questions about the API, the Agent or integrations, write to developers@prava.dev. For active incidents, contact soc@prava.dev.
Note. This page is an overview. Full documentation, including the complete API reference, is available to customers inside the platform.
Next step
Start building with Prava AI
Talk to us about your deployment, or create a workspace and follow the steps above.