👉 Try the Live Interactive Prototype Here
Transforming Perpetual KYC from manual audits into AI-augmented, real-time risk resolution.
pkyc-video1.webm
1. The Landscape & The Problem
In the regulatory space, onboarding (KYC) is only the beginning. Financial institutions are mandated to perform Perpetual KYC (P-KYC)—continuously monitoring existing users for changes in their risk profiles (e.g., being added to a PEP watchlist, sudden anomalous wire transfers).
The Enterprise Shortfall:
L1/L2 Compliance Analysts and MLROs are currently drowning in false positives and fragmented data.
- Swivel-Chair Compliance: Analysts are forced to jump between 5 different tools (transaction logs, identity databases, watchlists) just to piece together why an alert was triggered.
- Missed SLAs: High-volume alerts lacking contextual identity data lead to slow response times, resulting in severe regulatory fines.
- Manual Overhead: Resolving a single alert requires analysts to manually draft complex Suspicious Activity Reports (SARs) or Request for Information (RFI) emails.
2. The Hypothesis
What if we could centralize the fragmented data into a single "Living Profile" and use Generative AI to do the heavy analytical lifting? If we aggregate transactional data and identity profiles into one triage queue, and pair it with an AI Copilot that reads the timeline to summarize the risk in plain English, we can drastically reduce the Time-to-Decision (TTD) on critical alerts.
3. The Solution & UX Strategy
I designed the P-KYC Engine as a high-density, single-page application specifically tailored for the cognitive load of compliance officers. Focused on the two most critical screens: The Triage Queue and the Living Profile.
Key Design Decisions:
- SLA-Driven Triage Queue: The global dashboard prioritizes urgency. The data table surfaces the exact Alert Trigger, the Current Risk Score (with trend indicators), and a prominent visual SLA countdown timer so analysts know exactly what to tackle first.
- The "Living Profile": When reviewing a flagged entity, the UI utilizes a strict 70/30 split layout. The left column houses the irrefutable data: a circular Risk Gauge, prioritized Risk Drivers (e.g., PEP Match vs. Wire Transfer), and a chronological Activity Timeline.
- Contextual AI Copilot: Instead of a generic chatbot, the right column houses a sticky AI Resolution Copilot. It continuously reads the left column's timeline and generates a concise, one-paragraph analytic summary of why the user exceeded the institution's risk appetite.