Easy AI · Workload brief
(Attach to an email to yi__ai@126.com. Write "open" where unsure.)

1. Basics
Company or team:
Contact and role:

2. Workload
Type (inference / parameter-efficient fine-tuning / full training):
Purpose:
Model, size and precision:
Concurrency and context length (inference):
Dataset size and checkpoints to keep (training):

3. Resources
Compute (must have / substitutable):
Storage (volume, growth, backup needs):
Network (who connects, bandwidth, isolation, fixed addresses):

4. Timing
Start date:
Duration:
Interruptible periods and maintenance windows:
Exit date:

5. Responsibilities
We handle:
We would like you to handle:
To agree:

6. Data and compliance
Data classification and sensitivity:
May data be stored in a partner data center:

Do not put admin passwords or real business data in this brief.