> ## Documentation Index
> Fetch the complete documentation index at: https://docs.avigrah.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Refund Policy

All payments made to Avigrah are **non-refundable**.

Once services are provisioned, infrastructure is allocated, and execution capacity is reserved for the organization, costs are incurred irrespective of actual usage.

This applies to:

* Subscription commitments
* Setup and onboarding fees (if applicable)
* Usage allocations (Intelligent Units, storage, execution capacity)

Organizations are encouraged to evaluate requirements during onboarding to ensure alignment with expected usage.

***

### Usage Validity & Allocation Period

All allocated usage, including:

* Storage (data, vector, lakehouse)
* Intelligent Units
* Execution capacity (tasks, automations)
* Infrastructure limits

must be utilized within the **active contract period**.

Avigrah does **not operate on a credit-based or rollover system**.

Unused capacity:

* Does not carry forward to the next billing cycle
* Expires at the end of the commitment period (3 / 6 / 12 months)

***

### Why No Rollover?

Avigrah provisions dedicated infrastructure and execution capacity based on your agreed usage.

This includes:

* Reserved compute for AI workloads
* Storage allocation across multiple layers
* Continuous system availability for real-time and batch execution

As a result, capacity is **reserved, not on-demand**, and cannot be reallocated retroactively.

***

### Flexibility & Adjustments

If your requirements change during the contract period:

* Usage limits can be **scaled up**
* Additional capacity can be provisioned
* Contract terms may be revised based on new requirements

All changes are handled through mutual agreement and updated commercial terms.

***

### Key Principle

Avigrah is designed for **predictable, production-grade AI execution**, not pay-as-you-go experimentation.

* Capacity is planned upfront
* Infrastructure is reserved
* Execution is guaranteed

This ensures reliability, performance, and governance at scale.