Skip to main content
ORDRANISecurity

For cloud security teams

See the automation inside your cloud — not only the API calls it makes.

Ordrani gives cloud security teams a behavioral view of the roles and workload identities operating across their AWS accounts — what each one normally does, which automation is responsible, and when an authorized identity steps outside its purpose.

The challenge

Cloud logs record what happened. They rarely say why.

Cloud audit logs record what happened in exhaustive detail, yet say little about why. A deployment role updating a service and the same role reading a production secret appear as two well-formed, authenticated events. Telling them apart requires knowing what the role exists to do.

As organizations spread workloads across many accounts and regions, the number of roles, trust relationships and cross-account paths grows faster than any team can reason about manually.

How Ordrani helps

  • Behavioral baselines for every monitored role and workload identity across accounts
  • Detection of roles exercising dormant permissions or reaching new accounts and regions
  • Context on which workflow, repository or service is responsible for cloud activity
  • Investigations that connect control-plane changes to the identities and executions behind them

Signals evaluated

  • AssumeRole chains
  • new region or account
  • dormant permission use
  • trust policy change
  • data-event access

Who it serves

Cloud security leads and platform security engineers responsible for AWS environments.

Activity as a connected graph

Sessions, roles and resources, related over time.

Ordrani relates executions, sessions, operations and resources so that a sequence spanning accounts reads as one story.

Temporal identity graph

Every edge carries its time and the strength of its evidence.

OIDC claimssession ARNWorkflow rundeploy.yml · #4182Role sessionprod-deploy-roleecs:UpdateServiceexpected operationECS servicepayments-apiiam:ListRoles ×37novel for identityGetSecretValue4 min after runSecretprod/db-credentialsToken reuse?hypothesis only14:32:0514:32:0614:32:41 – 14:36:1214:36:12event time (UTC) →
Confirmed attributionStrong correlationUnverified hypothesis

How it works in AWS

From read-only telemetry to a cloud investigation.

  1. 01

    Connect read-only

    Read CloudTrail management and data events and IAM configuration through scoped, read-only access across your accounts.

  2. 02

    Model each role

    Build a behavior model for every monitored role and workload identity — what it calls, where, when and on which resources.

  3. 03

    Attribute activity

    Link role sessions to the workflows and services responsible, so activity is judged against its purpose.

  4. 04

    Raise cases

    Correlate deviations across accounts and regions into investigations with stated confidence.

What it detects

Cloud activity that looks authorized — and is not expected.

01

Dormant permission use

A role exercising permissions it has held for months but never needed.

02

New account or region

Automation reaching accounts or regions outside its established footprint.

03

Unusual role chaining

AssumeRole paths an identity has never followed, especially into more privileged roles.

04

Trust policy changes

Changes to who can assume a role, made by an identity that does not normally manage IAM.

05

Data-event anomalies

Access to objects, secrets or keys outside an identity’s normal scope, where data events are enabled.

06

Logging interference

Attempts by automated identities to stop, delete or reconfigure audit logging.

Bring behavioral context to your cloud accounts.

We would welcome a technical conversation about the roles, accounts and automation in your AWS environment.