Schedules

Schedules allow AI systems to run automatically without requiring manual intervention.

They enable organizations to execute AI workflows at predefined times, recurring intervals, or business-specific schedules.

This transforms AI from a reactive tool into a proactive operational system.

Why Schedules Exist

Many business processes occur on predictable timelines rather than in response to user requests.

Examples include:

Daily reports
Weekly summaries
Monthly compliance checks
Morning standup generation
Customer health monitoring
Security scans
Cost optimization reviews

Schedules automate these repetitive activities.

How Schedules Work

A Schedule acts as a trigger.

At the configured time, Karyam automatically starts the associated AI Flow.

Schedule
    ↓
AI Flow
    ↓
Skills / MCP Tools
    ↓
Output

The resulting execution appears as a standard Run and participates in all observability and governance features.

Common Use Cases

Organizations commonly use schedules for:

Reporting

Daily sales reports
Weekly engineering summaries
Monthly executive dashboards
Customer engagement reports

Monitoring

Infrastructure health checks
Security reviews
Cost analysis
Dependency monitoring

Automation

Backup verification
Data synchronization
Ticket generation
System cleanup tasks

AI Operations

Knowledge synchronization
Content generation
Research updates
Scheduled summarization

Schedule Configuration

Schedules typically define:

Name

A human-readable identifier for the scheduled task.

Example:

Daily Sales Summary

AI Flow

The workflow that will execute when the schedule triggers.

Examples:

Daily Reporting Flow
Customer Health Check
Security Audit Workflow

Frequency

Defines how often the schedule executes.

Examples include:

Hourly
Daily
Weekly
Monthly
Custom CRON expressions

Timezone

Schedules execute relative to the selected timezone.

This ensures workflows trigger at expected business hours regardless of infrastructure location.

Example:

Asia/Kolkata
UTC
America/New_York
Europe/London

Example

Every day at 08:00
        ↓
Generate Daily Operations Report
        ↓
Email Results to Leadership Team

No manual intervention is required.

Scheduled Runs

Scheduled executions create normal Runs.

This means scheduled workflows automatically benefit from:

Logs
Token tracking
Cost analytics
Audit trails
Approval workflows

From an observability perspective, scheduled runs behave exactly like manually triggered runs.

Schedules and Approvals

Schedules do not bypass governance controls.

If a scheduled workflow requires approval:

Schedule
    ↓
Workflow Starts
    ↓
Approval Required
    ↓
Wait for Approval
    ↓
Continue Execution

Human approval remains mandatory where configured.

Schedules and Failures

If a scheduled execution fails:

The Run is marked as failed.
Logs are captured.
Audit events are generated.
The next schedule execution proceeds normally.

Failures do not automatically stop future executions.

Relationship to Other Concepts

Schedule
    ↓
AI Flow
    ↓
Run
    ↓
Logs

Schedules provide the automation layer that allows AI systems to operate continuously without human initiation.

The Karyam Philosophy

AI systems should not only respond to requests.

They should proactively perform work on behalf of the organization.

Schedules make AI systems continuously operational.