Data Engineer & GenAI Finance Specialist

Lê Phúc Chí

+84-352-***-*** ***********@*****.*** GitHub LinkedIn

Education & Awards

University of Economics and Law – VNU-HCM Ho Chi Minh City, Vietnam Bachelor of FinTech – Finance and Technology Expected 2026

• 1st Prize – UEL Young Researcher Award 2026 (University-level Final, Scientific Research)

• Top 12 – The Finnovative Hackathon

Experience

Ernst & Young (EY) Ho Chi Minh City, Vietnam

Data Engineer Dec 2025 – Jul 2026

• Executed end-to-end data migration from Oracle EBS R12 (20+ year legacy system) to Oracle AI Database 26ai: designed and implemented Extract, Transform, and Load pipelines tailored to active business operation requirements across functional modules; extracted and validated source data using PL/SQL scripts and custom data packages; transformed data to align with target schemas and business rules; loaded into Oracle AI Database 26ai via WebADI and SQL Loader invoked through Oracle Concurrent Programs, ensuring data integrity and continuity across GL, AR, AP, and fixed asset modules.

• Researched and demonstrated enterprise lakehouse architecture on Microsoft Fabric and Azure Databricks, exploring OneLake storage and Medallion design patterns (Bronze: raw ingestion, Silver: PySpark/Databricks transformations, Gold: curated analytical layers); prototyped batch ETL workflows via Fabric Data Factory to evaluate processing of multi-source financial datasets into analytical layers.

• Researched and developed GenAI systems in the data domain leveraging the Azure AI Foundry ecosystem – integrated LLMs for NL2SQL and enterprise knowledge retrieval; containerized and deployed inference services on Azure Kubernetes Service

(AKS) as scalable microservice endpoints.

Projects

Deep Learning Signal & RAG Investment Advisory Platform 2025–2026

• Built dual-pipeline batch ETL (Airflow-orchestrated) on AWS Medallion lakehouse (Bronze/S3 Silver/Glue: 43 technical features + Triple Barrier labeling Gold/Athena) for 27 Vietnamese banking stocks (106,650 OHLCV records, 2010–2026); parallel knowledge pipeline indexes 11,500 financial document chunks into 3 isolated HNSW vector indices.

• Engineered CNN-LSTM-Transformer (CLT) for 3-class signal classification (Buy/Hold/Sell) over a 60-day lookback; achieved 56.2% accuracy and F1-Macro 0.431 under Purged Walk-Forward CV, outperforming LSTM, Random Forest, and Logistic Regression baselines.

• Designed SR-MI-RAG advisory system: LLM semantic router (τ=0.32) dispatches queries to correct HNSW index; BGE-M3 embeddings (1024-dim) retrieve top-K=5 passages, achieving 87.3% source-grounded responses and 98.6% routing accuracy. Deployed via FastAPI/Docker; backtested Sharpe Ratio 0.233 and 47.7% MDD reduction vs. Buy & Hold. Multi-Index RAG Platform for Vietnamese Finance & Law 2025

• Architected a multi-index RAG pipeline (LangGraph) over 1.5M+ documents across 4 specialized pgvector indices (Legal 15K, Financial 1M, News 500K, Glossary 3K) on Supabase; hybrid semantic + rule-based router achieves 100% routing accuracy at 5ms with automatic query decomposition for complex multi-intent queries.

• Built AI chat service with streaming SSE, 3-tier Redis cache, QueryGuard prompt-injection defense, and web search fallback; integrated BGE-M3 embeddings with Gemini 2.0 Flash (primary) and GPT-4o (fallback). Delivered full-stack product

(Next.js 15 + FastAPI) with OAuth 2.0, Swipe News, Portfolio Tracker, Docker/Nginx, end-to-end <500ms. Real-Time Market Simulation Platform 2024–2025

• Built real-time trading simulator with WebSocket price streaming and order execution simulation; containerized with Docker/NGINX and CI/CD pipelines for scalable cloud deployment. Technical Skills

Data Platform & Engineering: PySpark, Airflow, Medallion Lakehouse, Microsoft Fabric, Azure Databricks, Azure Data Factory, AWS (S3, Glue, Athena), Oracle Development, PL/SQL (Window Functions, CTEs, Stored Procedures, Data Packages)

AI & Machine Learning: PyTorch, TensorFlow, Scikit-learn, LangGraph, RAG Systems, HNSW/FAISS, BGE-M3, Azure AI Foundry, Azure OpenAI

Cloud & DevOps: Docker, Amazon Web Services, Azure, CI/CD, NGINX Software & Analytics: FastAPI, Next.js, REST APIs, WebSocket, Redis, Power BI, Git, Oracle EBS Publication

Le Phuc Chi et al. “Optimizing Car Rental Services Using Smart Contracts on Blockchain Platform.” Tap Chi Cong Thuong, 2025. doi.tapchicongthuong.vn/nckh/article/view/72 Certifications & Language

Technical: NVIDIA DLI – Fundamentals of Deep Learning (2025) Data Engineering Bootcamp (2025) Language: VSTEP B2 – Level 4 TOEIC 745/990

Contact this candidate