Shawn Ng

DBS, a leading financial services group headquartered in Singapore, operates across 18 markets. We are an Asia-centric commercial bank focused on harnessing the region's long-term potential as the centre of economic gravity shifts eastwards to Asia

Key Achievements: Architecting the Enterprise AiOps Engine

• Spearheaded the end-to-end research and implementation of a next-generation AiOps engine (LLMs/ML/statistical models), reducing Mean Time to Resolution (MTTR) from hours to minutes (1-10 min) through proactive anomaly detection.
• Built robust data & CI/CD pipelines integrating enterprise monitoring tools (Zabbix/Sonar) to automate the testing and deployment of AI models.
• Optimized AI performance & resilience by developing a tracking framework for model stability and upstream data quality, significantly lowering operational costs.
• Directed stakeholder alignment by benchmarking AI performance against human analysts, identifying data gaps, and setting realistic KPIs for cross-functional leadership.

Core Infrastructure & Leadership

• Directed a team of 5 developers, aligning the multi-year product roadmap with group-level business objectives while managing stakeholder expectations across product, security, and compliance.
• Architected self-service DBaaS platforms (Neo4j, SingleStore), automating security-hardened deployments that reduced provisioning time by 98.2%, accelerating time-to-market for 12,000+ developers.
• Engineered 99.9% uptime via an automated health-monitoring ecosystem utilizing Prometheus, Grafana, and Python auto-healing scripts.
• Modernized legacy data infrastructure by consolidating services across Kafka, Python, and Node.js, drastically reducing technical debt.
• Automated security-compliant Hadoop deployments, ensuring rigorous adherence to financial regulatory standards.
• Expanded data product capabilities by maintaining and developing six Python Flask apps (encryption, file routing, database connection, Neo4j, Prometheus).
• Enhanced server health monitoring by developing an NLP model that analyzes logs (Graylog, MariaDB), providing actionable insights and recommendations.

Seed round; Urbanzoom is a high tech startup that utilizes artificial intelligence to value real estate properties and match homeowners or owners-to-be to the ideal real estate agent

• Orchestrated high-stakes POCs with OneMap, DBS, and OCBC, designing API integrations that validated the core product and secured strategic corporate partnerships.
• Architected a scalable hybrid-cloud infrastructure (AWS/GCP), managing S3 data lakes, BigQuery warehouses, and RDS databases to support enterprise-level operations.
• Pioneered a 3D data acquisition strategy by managing vendor contracts for drone-based aerial photogrammetry, integrating high-fidelity spatial models into the property valuation engine.
• Engineered ETL pipelines (Airflow) to ingest and process 10TB+ of spatial-temporal datasets, automating the scraping and cleaning of unstructured data to boost team productivity.
• Productized 10+ APIs and ML models (XGBoost, Scikit-learn) for automated property valuation and footfall prediction, directly enabling new revenue streams and customer-facing features.
• Maintained full-stack production codebases across Django, Express, and Ruby on Rails, ensuring seamless integration between data models and the web application.

Series A; Regtech company that uses big data analytics and cutting edge technology to revolutionise the way companies protect themselves from criminals, terrorists and money launderers

• Spearheaded production deployment for Standard Chartered Bank and managed POCs for Aviva, DBS, and Bank of Singapore, directly expanding the company's enterprise footprint.
• Engineered an NLP-based feature extraction engine using SpaCy and TextBlob, boosting ML model accuracy by 20% through optimized text preprocessing and standardization pipelines.
• Achieved 100% recall in fraud detection by architecting a robust outlier detection and filtering system, effectively eliminating false positives while maintaining strict adherence to complex business rules.
• Developed scalable data pipelines utilizing Python (Pandas, PySpark), SQL, and Docker to clean and ingest multi-source data for high-frequency ML model training.
• Bridged the gap between business and engineering by translating complex AML requirements (Name Screening, Transaction Monitoring) into technical specifications for development teams.

Seed round; Y-Combinator S14 fintech startup that makes it easy for users to spend digital currencies such as bitcoin

• Increased staff productivity in analyzing users' transaction behavior, fraud detection, and disputed transactions by 100% by building 4 visualization dashboards (D3.js) and 12 reports on the web app.
• Increased user experience by 100% by increasing fund transfer and user information update speed by building an admin dashboard (Angular, Ruby on Rails) features.
• Increased customer satisfaction by resolving 5,000+ customer technical issues and collecting product feedback.
• Increased accounting transparency by analyzing 0.5M transactions and VISA invoices (ISO 8583).