Skip to content

Home

Jump to bottom
martapiekarska edited this page Oct 9, 2024 · 1 revision

FIDL Enterprise Allocator Policies

Data Ownership

  • FIDL Enterprise Allocator only assigns Data Cap to clients who are owners of the data.
  • The client and Storage Provider cannot be related - the KYB process will be focused on verification of true separation of the two entities.
  • As part of the audit process, FIDL Enterprise Allocator team will request proof of payments between client and the Storage Providers. To provide privacy to client and SP we expect to be able to review an audit from an approved third party auditor.

Formatting of Stored Data

  • There must be clear documentation provided on how the data is prepared / transformed between the source data set and the data stored in Filecoin
  • Deals / Pieces stored to Filecoin should be full - there should not be significant 'slack' or 'padding' in stored data

Usage of DataCap

  • There is an expiration date of three months on any allocation of DataCap. From an allocation date, we will measure three months time and if the allocation has not been used (open or closed status), the application will be closed and remaining DataCap removed.
  • The expectation when the complete amount of requested DataCap is allocated is that the client has completely finished onboarding their dataset and replicas. If a client receives a DataCap allocation, then closes their application before completion, they will be questioned as to why. Likewise, if a client receives a DataCap allocation and abandons the application and becomes completely non responsive, their GitHub ID will be flagged from any future participation in the pathway.

Process

Our diligence efforts focus on confirming:

  • the GitHub applicant user (KYC)
  • the Data Owner business (KYB)
  • the Dataset (contents and size)

Clients and data owners are vetted manually upfront to confirm who they are, what data they are onboarding, how they will prepare the data and which SPs will be involved in onboarding copies of the data.

Application

All responses are reviewed and assessed by FIDL team based on the following:

  • Dataset - validation of data to be onboarded. Because the data is private encrypted data, the client is required to show a manual demonstration of the data (share raw data or screen share) with allocator team.
  • Data Owner - confirmation of connection to business requesting onboarding.
  • Client - confirmation of applicant and connection to dataset.
  • Data Preparation - who is preparing, what tool(s) are used.
  • Retrievability - what are the requirements of the Data Owner in terms of data retrieval.
  • GitHub ID - Due diligence involves confirming usage associated with the client GitHub ID to enable applicants to add a layer of trust and ultimately utilize one GitHub ID and build a GitHub ID reputation as a good actor over time.

New User Check

The first checks to be completed on each application by the allocator:

  • Is this a completely new GitHub ID? (less than 2 months old)
  • Is this the first time this GitHub ID has applied for DataCap in this or other pathways?
    • If yes to either, applicants will have a lowered maximum DataCap allowance.

Client Check (KYC)

Because the contents of the dataset are not publicly retrievable, we require all clients to complete a know your customer, (KYC), check to confirm themself as a human user, through https://kyc.allocator.tech/ Clients can also choose another method to prove the identity of the applicant (which must first be vetted by the allocator team and made public for transparency).

Business Check (KYB)

Because the contents of the dataset are not publicly retrievable, all dataset owner businesses must complete a know your business, KYB, check to confirm the business exists, is legit, and someone from the data owner team (business) has approved onboarding of the dataset. We currently offer two options for KYB check:

  • We offer a third party KYB service partnership with Synaps.io ($100), that will review business and data owner documentation (details in answer https://github.com/filecoin-project/notary-governance/issues/18 below)
  • Or they can plan a virtual meeting with the client, data owner and member of the allocator team to review:
    • the dataset
    • confirm ownership (proof of employment, employer signoff, sharing the business license).
    • and validate storage of the data by the client/applicant is approved and a contract is in place

The main verification will focus on ensuring that Data Owner and the Storage Provider are not colluding. FIDL allocator team will request questions to ensure that the two entities are entering a purely business engagement and not collaborating to extract DataCap

If requested by a client, we will utilize various non-disclosure agreements to collect required information on clients and data-owners while maintaining their privacy.

Clone this wiki locally