01 — Context
BRISPOT Pemutus is the credit decision counterpart to the Pemrakarsa application. Where Pemrakarsa handles loan proposal initiation by Relationship Managers, Pemutus is used by credit approval officers to review completed proposals, evaluate credit risk analysis, and make final approval decisions.
Because Pemutus is the point where financial commitments are made, accuracy and reliability are the defining constraints. The application presents complex credit analysis data clearly and completely, ensuring approvers have full context before acting.
02 — Role
Android Developer. Contributed to feature development, security improvements, and ongoing maintenance of the application, working within a cross-functional Agile team.
03 — Engineering
The primary engineering challenge in Pemutus is presenting dense, multi-dimensional financial data clearly and reliably. Approval officers work under time pressure with many proposals to review — the UI must surface critical information quickly without losing completeness.
Architecture
- Clean Architecture + MVVM: Same disciplined layer separation as the broader BRISPOT platform. Domain layer handles credit evaluation business rules independently of any UI concern.
- Data Layer: Repository pattern abstracts the REST API communication behind clean interfaces. The UI is never aware of whether data came from a network call or a local cache.
Security
- Retrofit + Certificate Pinning: Encrypted communication with core banking APIs. Auth token management and secure request headers are handled via OkHttp interceptors.
- Compliance: Features are designed with banking regulatory requirements in mind. Data presentation follows strict accuracy requirements.
Reliability
- Room: Local caching ensures approvers can review previously loaded proposal data even in poor network conditions, without data inconsistency.
04 — Engineering Challenges
Presenting complex multi-source data with guaranteed accuracy.
A credit decision involves pulling data from multiple backend systems: proposal data, customer credit history, collateral valuations, and compliance checklists. Reconciling these into a coherent, trustworthy view — and handling partial or degraded data gracefully — is the core UI/data problem.
Preventing data staleness in a high-stakes context.
In a credit decision context, displaying outdated data is not a UX inconvenience — it is a risk. Cache invalidation strategies and explicit data freshness indicators are part of the engineering design.
05 — Contributions
- Developed features for the credit review and approval workflow
- Implemented secure API integration for credit data retrieval
- Worked on local data caching and cache invalidation strategies
- Collaborated with compliance and QA teams to meet banking standards
- Contributed to ongoing maintenance and stability improvements
06 — Technologies
KOTLIN Primary language
MVVM Presentation architecture
CLEAN ARCH Layer separation
RETROFIT Secure REST API networking
ROOM Local caching / data reliability
COROUTINES Async data loading
FLOW Reactive UI state