Questions and Answers for SAM Revised.docx

November 7, 2024

Questions and Answers #1

1. Could you provide the current amount of storage space in use?

ANSWER: 260GB

2. What are the anticipated future storage requirements? For example, would an estimate of 1TB, 2TB, or more?

ANSWER: Est. 1T

3. How many users are expected to access the network?

ANSWER: 25 users are expected.

4. Could you confirm if Google Cloud Platform (GCP) would be an acceptable alternative to Azure for hosting a DAM solution?

ANSWER: Must be on our Azure server.

5. Is the USDA open to using a commercially available Azure-based DAM solution, or does it prefer a custom-built system?

ANSWER: Custom-built system to mimic as closely as possible the look and functionality of our existing Asset Bank DAM.

6. What specific challenges or limitations has USDA encountered with the current Bright Interactive Asset Bank DAM?

ANSWER: We are extremely happy with Asset Bank and would not leave it except for the fact that they will no longer support a system hosted on our server. Over the years they were able to accommodate all our requests for new functionalities except one – the global lookup function described in 3.1.x.

7. Are there specific engineering or implementation tasks that must be performed onsite, or can most tasks be managed remotely?

ANSWER: Can be managed remotely.

8. Is there a preferred approach or tool for migrating the 46,000+ digital assets and associated metadata from the current DAM to a new system?

ANSWER: No

9. Can USDA provide a sample of current metadata structures and content types to test feasibility?

ANSWER: We can provide screen shots of Asset Bank or download a 240MB SQL file for the entire database.

10. Are there any specific data validation or testing criteria that USDA requires during migration?

ANSWER: No

11. Given the project’s Azure environment requirement, does USDA have a preference between an Azure-native solution, such as MediaValet or Bynder, versus a custom DAM built on Azure resources like Azure Blob Storage and Azure SQL?

ANSWER: A custom DAM built on Azure resources like Azure Blob Storage and Azure SQL

12. For optimizing scalability and performance in the Azure Cloud, are there specific architectural requirements from USDA, such as geographic redundancy, containerization, or compliance with FedRAMP for government cloud environments?

ANSWER: It does not have to be because we are going to be using USDA storage so it does not have to be fed ramp.

13. Could you share any insights into the types of workloads (e.g., retrieval frequency, bulk downloads) anticipated for the DAM to help us better gauge the cloud infrastructure needs for peak demand scenarios?

ANSWER: This is a database that will be used by less than 5 people within our organization at the same time. Up to 25 people will have permission to login to search, upload, and download assets. We don’t usually have large workloads.

14. The PWS mentions hierarchical controlled vocabularies and customizable metadata fields. Are there specific taxonomies or metadata schemas currently in place that must be carried over, or is the USDA open to adopting an enhanced structure aligned with industry standards, such as IPTC for digital assets?

ANSWER: YES - there are specific taxonomies or metadata schemas currently in place that must be carried over. See attached samples.

15. Does USDA foresee the need for advanced metadata interoperability with other systems, such as integration with external repositories or databases, or even future integration with machine learning models for automatic metadata tagging?

ANSWER: We do not have any current integration with other databases, but it would be good to have the option for the future. We prefer not to have any AI or automatic metadata tagging.

16. Considering the scale of migrating 46,000+ assets, can USDA provide insight into the data quality and structure of current assets, such as common metadata fields, file naming conventions, or any existing data anomalies that may affect migration?

ANSWER: Can provide the following: a 240MB .SQL file of the entire database or screen captures.

17. Could you elaborate on the preferred approach for migration verification and integrity assurance? For instance, would a sample-based audit approach suffice, or does USDA require comprehensive checks across all migrated assets?

ANSWER: The USNA staff will do spot checks and ample testing after the migration.

18. Is there a specific requirement for maintaining historical metadata or revision histories from the current system, particularly in terms of compliance or traceability needs?

ANSWER: Yes

19. The system must allow flexible user permissions for partitioned access to assets. Could you clarify if the permission model follows a role-based approach, and if so, how many distinct user roles USDA envisions? Are there anticipated custom roles that go beyond standard read, write, and admin permissions?

ANSWER: We want the ability to create and assign permissions.

20. Are there particular security protocols or frameworks USDA requires for authentication and access control, such as integration with USDA’s existing single sign-on (SSO) systems or multi-factor authentication requirements?

ANSWER: We require users to log into the DAM. Since they can only access it from a government-issued computer we do not need the USDA-SSO.

21. Regarding the integration with USDA’s Arboretum Botanical Explorer and mobile app, are there existing APIs or web services to support seamless DAM synchronization? If not, does USDA prefer a specific protocol (e.g., RESTful APIs, GraphQL) for enabling these connections?

ANSWER: There is none currently (it is done manually), but we may prefer to connect in the future.

22. The PWS mentions scriptable exports to remote servers for selected assets. Are there preferred automation tools or scripting languages (e.g., PowerShell, Python) USDA uses, or is the DAM provider expected to introduce a compatible solution?

ANSWER: DAM provider is expected to introduce a compatible solution.

23. For the duplicate detection feature, would USDA prefer using pixel-level duplicate detection algorithms like perceptual hashing, or is there interest in leveraging machine learning-based models for enhanced accuracy, especially for visual similarity searches?

ANSWER: Pixel-level duplicate detection.

24. Are there custom workflows or digital asset lifecycle stages USDA envisions, such as approval workflows for archival or public publishing, that would require a tailored approach?

ANSWER: TBD

25. Would USDA prefer proactive system health checks, such as automated alerts for system anomalies, or specific tools within the Azure ecosystem for monitoring and diagnostics (e.g., Azure Monitor, Log Analytics)?

ANSWER: Yes

26. Could you confirm any additional security protocols USDA requires in compliance with federal standards like FIPS 140-2 for cryptographic modules or specific NIST guidelines on data encryption and logging?

ANSWER: TBD

27. Does USDA have specific standards for logging access and changes to digital assets (e.g., detailed audit trails, tracking metadata edits), especially in high-sensitivity collections?

ANSWER: We want the ability to track metadata edits to an asset and audit trails for uses.

28. For user training and knowledge transfer, is there an expected format (e.g., in-person, virtual, recorded sessions) and duration USDA requires for effective end-user and admin onboarding?

ANSWER: Virtual training is preferred for onboarding users. Additional virtual training will be required for administrators.

29. Beyond standard support, does USDA envision specific Service Level Agreements (SLAs) for incident response times or escalation procedures that align with its internal IT policies?

ANSWER: No

30. Are there future scalability expectations for this DAM to expand to other USDA collections, and would USDA benefit from a modular architecture to facilitate future feature expansions?

ANSWER: No Plans.

31. Can you provide an example of the hierarchical controlled vocabulary that should be uploaded?

ANSWER: See examples provided.

Categories = drilldown category tree Photographer = dropdown list, click to choose or add new name in separate field Photo Location = keyword list arranged alphabetically and hierarchically in a drilldown list, click to choose from admin-controlled list Keywords = keyword list arranged alphabetically, click to choose or add new word in separate field Original format = dropdown list

32. We assume that each asset will have only one version. Is this assumption correct?

ANSWER: Yes

33. Can you please describe your existing workflows in detail, or provide a workflow diagram?

ANSWER: To obtain DAM data for ABE and the mobile app, we currently perform a MySQL data dump on the server that hosts the DAM (either a .sql or a .txt file), then transfer the resulting file to a local machine for further extraction and manipulation. Data from two of the current tables in the DAM’s database (one, called “assetattributevalues,” primarily contains metadata associated with each digital asset; the other, called “asset,” contains further data about each digital asset) are extracted to a .csv file, which undergoes additional human manipulation.

To obtain the images needed for use in ABE and the mobile app, we use built-in script features in the existing DAM to set up a scripted selection routine that creates several .jpg thumbnails for each digital asset, one thumbnail each for several different reduced sizes that we define (and can redefine), and then publishes those thumbnails to one of several folders on a remote server. One folder is used for each thumbnail size category, so that each folder contains a thumbnail version of every digital asset. The thumbnail files are named based on one or more criteria that we have specified; we use only the digital asset’s ID number. Thus, each thumbnail created for an original digital asset with ID number 12345 will be named 12345.jpg. There will be several different-sized files named 12345.jpg, one published to each of the different thumbnail folders on the remote server. After initially publishing of all digital asset thumbnails to those remote server folders, running the script (currently performed by manually clicking a button) will select and publish only those assets that are new or that have changed since the last publishing event.

34. How we access to the existing Bright Interactive’s Asset Bank DAM be made available? On site in Washington DC on USDA hardware, or will remote access be feasible?

ANSWER: Remote access may be provided.

35. How can we identify the ID number in the current Bright Interactive’s Asset Bank DAM?

ANSWER: Bright Interactive’s Asset Bank DAM hosted on a private server, version 3.1828.45.54.

36. If some information cannot be extracted using the API, it will not be extracted. Is this approach acceptable?

ANSWER: It is my understanding that we do not have the API function enabled in our version of Asset Bank.

37. Is it required to support a 1 to 1 attribute to metadata mapping between Bright Interactive’s Asset Bank DAM, or will some attributes be merged or even omitted from migration?

ANSWER: We require a one-to-one metadata mapping.

38. Cloud Hosting (Section 3.2 a): "Will the vendor be responsible for providing the cloud hosting infrastructure for the production system, or will the government provide the hosting environment?"

ANSWER: A government provided hosting environment.

39. Cybersecurity Authorization (Section 3.2 d): "Is the production system required to obtain a cybersecurity authorization-to-operate (ATO)? If so, will the vendor be responsible for obtaining the ATO, or will the government handle this process?"

ANSWER: No

40. Current System Database (Section 3.3 u): "What type of database is used in the current digital asset management system? Please provide details on the database software, version, and schema."

ANSWER: Bright Interactive’s Asset Bank DAM hosted on a private server, version 3.1828.45.54.

41. Database Connection (Section 3.3 u): "The PWS states that the new system must connect to the current system's database. However, it also requires the transfer of all digital assets to the new system. Can you please clarify the purpose of this database connection and how it relates to the transfer of assets?"

ANSWER: There is no requirement that it connect to the old DAM. The old DAM will be going away after all the assets and metadata have been successfully migrated to the new system, and testing shows that the needed capabilities and linkages have been replicated. Section 3.1.u refers to our apps that we populate with information exported from the DAM periodically.

42. Temporary Connection (Section 3.3 u): "Is the connection between the new system and the current system's database intended to be temporary? If so, what is the expected duration of this connection?"

ANSWER: No connection is required.

43. System Deployment (Section 3.2 a): "Will the vendor be responsible for deploying and configuring the new DAM system in the Azure Cloud environment, or will government personnel handle the implementation?"

ANSWER: The vendor is responsible for deploying and configuring the new DAM system in the Azure Cloud environment.

44. Asset Transfer (Section 3.2 c): "Will the vendor be responsible for developing and executing the migration script to transfer digital assets to the new system, or will government personnel manage this process?"

ANSWER: The vendor is responsible for developing and executing the migration script to transfer digital assets to the new system.

45. Production Deadline (Section 3.3 a): "What is the deadline for fully operationalizing the new DAM system in the production environment?"

ANSWER: Ideally by the end of February, but it is negotiable.

46. ATO Deadline (Section 3.2 d)?

ANSWER: No Needed

47. Asset Transfer Deadline (Section 4.0): "What is the deadline for completing the transfer of all existing digital assets to the new DAM system?"

ANSWER: Ideally by the end of January, but it is negotiable.

48. Cloud Hosting Costs (Section 3.2 a): "Should the vendor's price quote include the costs associated with hosting the DAM on the Azure Cloud? If so, please clarify whether this includes all hosting-related expenses, such as storage, computing, and network costs."

ANSWER: Vendor should not include hosting fees, storage, computing, or network costs in quote.

49. Workload Statistics (Section 3.2 a): "If hosting costs are included in the vendor's quote, could the government provide workload statistics to help estimate these costs accurately? This could include information such as the number of expected users, the estimated amount of data to be transferred per day/week/month, and any anticipated spikes in usage."

ANSWER: NA

50. Arboretum Botanical Explorer and the Arboretum mobile app migration to the new DAM (Section 1.0): “Could you please provide details on how the migration will be handled for the Arboretum Botanical Explorer and the Arboretum mobile app? How do the Arboretum Botanical Explorer and the Arboretum mobile app connect to the current DAM?”

ANSWER: There is no migration involving these two apps. The current DAM does not connect directly to the Arboretum Botanical Explorer or the mobile app. The asset thumbnails are created and published via a customizable script, but the data is obtained via a MySQL dump, and the dump file is then moved to a local machine for further extraction and manipulation of the needed data (currently two tables, “assetattributevalues” and “asset”).

51. The system has over 46,000 assets (Section 1.0): “Could you please provide the size of the current DAM database in gigabytes or terabytes?”

ANSWER: 260GB

52. IntelligenceBank offers its SaaS DAM on AWS US-West which is FedRamp compliant. It has been granted a Joint Authorization Board Provisional Authority-To- Operate (JAB P-ATO) and multiple Agency Authorizations (A-ATO) for moderate impact level. To greatly reduce annual costs, is USDA open to a multi-tenanted environment on FedRamp AWS-West? The instances (tenants) are logically isolated, but physically integrated. While the underlying hardware is shared, all customer servers are given exclusive access to dedicated memory and managed access to dedicated CPU and disk resources.

ANSWER: No.

53. How many TBs of data does your 46,000+ digital assets represent? Is that the amount of TBs you would like to quote?

ANSWER: Our digital assets are currently 260GB. We anticipate adding up to 1TB in the coming years.

54. How many Admin/Upload users to the DAM platform do you require?

ANSWER: 15 users

55. How many view and download users to the DAM platform do you require?

ANSWER: 25 users

56. Requirement S vs. X - can you please confirm that the "data lookup feature" isn't a must-have requirement right away, and one that can be built using our API for an additional cost once we have a better understanding of your requirements?

ANSWER: The data lookup feature (3.1.s) is required as part of the contract; the global lookup function (3.1.x) is a different feature that we would like to develop in the future.

Northeast Area Director’s Office 10300 Baltimore Avenue Building 003, Room 223 Beltsville, Maryland 20705-2350 USDA is an Equal Opportunity Provider and Employer