About

I spend my days at the intersection of data architecture and applied AI — designing the semantic foundations that make agentic systems actually work in production. 15 years across banking, government, and enterprise; most recently as VP, Data Analytics & Applied AI at Citi Singapore.

My background spans Accenture Singapore (7 years, government-embedded consulting across 10+ agencies), Citi (data engineering through AI leadership), and TCS. I've shipped production ML on ~2.5M citizens, designed enterprise data lakes, and built regulatory data quality engines from scratch.

Currently targeting forward-deployed AI engineering and solutions architect roles. Executive MBA, SP Jain · AI/ML Diploma, University of Hyderabad.

AI / GenAI / ML

  • LangGraph / LangChain
  • Mistral / OpenAI APIs
  • RAG Pipelines
  • NLP-to-SQL (Agentic)
  • Hugging Face

Data Platforms

  • Snowflake
  • Databricks
  • PySpark / Spark SQL
  • Hadoop / Hive / Impala
  • ChromaDB

Languages & Cloud

  • Python
  • SQL
  • AWS / Azure
  • Google Cloud

Architecture & Governance

  • BCBS 239
  • Solidatus / Collibra
  • Data Mesh Design
  • Agentic System Design

Experience

2020 —
2025

VP, Data Analytics & Applied AI

Citi Singapore

Led data analytics, applied AI, and governance across the bank. Shipped production agentic GenAI on regulated data, designed BCBS 239 data quality frameworks, and built a multi-agent NLP-to-SQL platform end-to-end.

LangGraphRAGSnowflakeBCBS 239Python

Agentic NLP-to-SQL + RAG Platform

2024 – 2025

Production multi-agent pipeline converting natural language to validated SQL, with a RAG layer over internal policy documents. Designed end-to-end as product owner and information architect on regulated banking data.

LangGraphMistral APIChromaDBSnowflakeDatabricks
—Users Served
—FTE Saved
—Query Time ↓

BCBS 239 Data Quality Engine

2021 – 2022

Sole architect and developer. Custom Python rules engine implementing BCBS 239 data quality controls — replaced a fragmented manual process with automated, auditable DQ validation across critical risk data domains.

PythonSQLCollibraSolidatus
—Data Domains
—Rules Automated
—FTE Saved
2013 —
2020

Manager, Data & AI Consulting

Accenture Singapore

Government-embedded consulting across 10+ Singapore agencies. Delivered production ML and enterprise data infrastructure at national scale.

PySparkHadoopReal-Time MLHive / Impala

Real-Time ML Fraud Detection

2016

Production ML fraud detection for a Singapore government agency covering ~2.5M citizens. Full pipeline — feature engineering, model training, and real-time scoring — delivered in a regulated government environment.

PySparkPythonHadoopStreaming
~2.5MCitizens Covered
—FTE Saved
Real-TimeDetection

Cross-Agency Government Data Lake

2017

14 TB data lake spanning 10 Singapore government agencies. End-to-end delivery including ingestion, transformation, and Tableau-ready query optimisation across heterogeneous data sources.

HadoopHiveImpalaPySparkTableau
10Agencies
14 TBData Lake
—FTE Saved
2011 —
2013

SAP Consultant

Tata Consultancy Services

Enterprise SAP consulting, delivering large-scale ERP implementations across multiple industries.

SAPEnterprise ERP

Open Source & Portfolio

● Work in Progress

SG Housing AI Platform

2025 · Personal Build · Being built in public

Multi-agent NLP-to-SQL and RAG system on Singapore HDB resale data. Architecture demo for agentic AI on structured public datasets — LangGraph orchestration, Mistral inference, ChromaDB for semantic retrieval, deploying to Hugging Face Spaces.

LangGraphRAGMistralChromaDBHugging Face

What's next?

Get In Touch

Actively exploring forward-deployed AI engineering and solutions architect roles in Singapore and beyond. If you're working at the intersection of data architecture and applied AI — or just want to connect — my inbox is open.

Say Hello