DEV Community

Query Filter
Query Filter

Posted on

gradle-103

MEMORANDUM

TO: Management

FROM: Leonid Spektor

DATE: September 1, 2026

SUBJECT: Architectural Assessment & Trade-off Analysis: Citi Desktop Onboarding & Atom Group Provisioning


Executive Summary

Onboarding our application suite to Citi Desktop introduces a structural constraint: the platform supports only one Atom Group per registered application in AppDirectory.

To map our existing 14 production user groups (~767 users across North America) into this paradigm, we have identified two viable onboarding strategies. Both options introduce operational trade-offs between application metadata management and ongoing access governance. Management decision is required to select the path forward.


Key Technical Constraint

Citi Desktop limits each AppDirectory application entry to a single bound Atom Group. Because our current access model relies on 14 distinct functional production groups (e.g., Group A, Group B, Group C), a direct 1:1 mapping within a single AppDirectory application entry is not natively supported by the platform.


Evaluation of Onboarding Options

Option 1: Application-Level Segmentation (14 AppDirectory Applications)

Create 14 separate AppDirectory application entries, where each entry differs solely by its associated Atom Group binding.

  • Pros:
  • Preserves all 14 existing functional Atom groups as-is without re-provisioning users.
  • Eliminates manual user re-keying and bulk entitlement migration risks.
  • Avoids ongoing dual-maintenance of user group memberships.

  • Cons:

  • High application clutter within the AppDirectory catalog.

  • Violates Atom team platform recommendations.

  • Increases administrative overhead during global configuration updates or application releases across 14 entries.


Option 2: Group Consolidation (2 AppDirectory Applications: Pilot & Prod)

Consolidate the existing 14 groups into 2 newly provisioned Atom Groups ([PILOT] and [PROD]), mapping to 2 corresponding AppDirectory application entries. This mirrors the approach currently being trialed by the EMEA team.

  • Pros:
  • Aligns with the Atom architecture team’s recommended deployment pattern.
  • Keeps the AppDirectory catalog clean and manageable (2 entries instead of 14).

  • Cons:

  • Initial Migration Risk: Requires mapping ~767 users into the new consolidated groups. Manual entry carries a high error rate and will require a dedicated QA validation cycle.

  • Long-Term Operational Overhead: Future user lifecycle events (onboarding, offboarding, role changes) will require updates across both legacy functional groups and the new Desktop groups.

  • Redundancy: Legacy functional groups must still be maintained if relied upon outside of Citi Desktop.


Comparative Trade-off Matrix

Metric Option 1: 14 AppDirectory Apps Option 2: Consolidated Pilot/Prod Groups
Catalog Maintenance High (14 app records to maintain) Low (2 app records to maintain)
Initial Implementation Effort Low (configuration-only in AppD) High (user mapping, entitlement migration, QA)
Data Entry / Error Risk Low (no user migration) High (~767 users to map without automated tooling)
Ongoing User Lifecycle Support Low (uses existing user entitlements) High (dual-group maintenance upon joins/moves/leaves)
Atom Architectural Alignment Non-Standard Standard

Recommendation & Next Steps

  1. Short-Term Priority: If Option 2 is selected, we strongly advise against manual user entry. We should engage the Atom support team immediately to request programmatic API or bulk CSV/scripting mechanisms to migrate the 767 users cleanly and reduce QA verification time.
  2. Decision Required: Management selection between accepting Application Catalog Maintenance (Option 1) versus Ongoing Access Maintenance & Migration QA (Option 2).

Top comments (0)