Data Scientist / Machine Learning Engineer with 2+ years building production fraud detection and anti-money laundering systems for 30+ bank clients — spanning anomaly detection, time-series forecasting, graph-based investigation, and agentic AI alert triage.
Sole ML owner across two divisions, from model development through client onboarding and deployment support. Now targeting AI Engineer roles building applied, production-grade agentic and LLM systems.
Sole ML engineer across Fraud Detection and Anti-Money Laundering divisions, owning the full model lifecycle — development, testing, client onboarding, deployment support, and maintenance — for 36 bank clients.
Identified pharmaceutical products with high market-growth potential using scraped pricing and sales data. Analyzed market-share across online and offline channels to guide product strategy. Built an advertising efficiency evaluation system to help optimize ad spend.
GPA: 3.46
Isolation Forest anomaly-detection on RFM-derived behavioral features with EWMA-based drift adaptation, processing 200K–800K transactions/day across 25 bank clients in near-real-time batch architecture.
Streaming anomaly-detection algorithm using random sparse projections with online histogram density estimation and EWMA-based adaptive thresholds — sub-200ms scoring at 133 TPS via Kafka/Flink.
Agentic AI system (LangGraph + AWS Bedrock) that investigates and explains flagged alerts to cut false-positive noise and analyst fatigue, reaching 93% agreement with analyst decisions across 25 clients.
Prophet-based time-series anomaly monitoring on transaction volume/frequency windows (200K–800K txn/day, 6 clients), paired with an agentic investigation pipeline and real-time WhatsApp alerting.
Graph-based detection system tracing multi-hop money-movement chains (up to 3 hops/day) across ~11K employees and ~90K daily transactions for suspicious transaction reporting.
A train service website offering a comprehensive and seamless experience for users seeking efficient and reliable train travel booking.
Attended the Huawei Intelligent Finance Summit in Shanghai, where Huawei unveiled upgraded solutions to drive finance toward agentic banking — accelerating the migration of financial institutions toward fully autonomous agentic banking systems.
.jpg)
Selected for a 24-day on-site engagement in the Philippines diagnosing an insurance client's customer-onboarding backlog — uncovering bottlenecks from legacy systems, failed OCR, and manual processes. Findings led to a follow-on core-insurance systems engagement.
Participated in the AWS AI League at AWS Summit Jakarta, a competition combining hands-on learning with AWS AI services (SageMaker AI & Amazon Bedrock). Competed against 30+ companies to build an AI solution for tourism and ranked in the top 10.
Interested in working together? Feel free to reach out.
