Ly Bao Tran Hoang

Honor Bachelor in Computer Science

August 2021 - January 2025

GPA: 8.92 / 10

Relevant Coursework

• Natural Language Processing
• Deep Learning
• Database Systems
• OOP
• Software Development
• Data Structure And Algorithm
• Operating Systems
• Computer Networks
• Machine Learning
• Computer Vision
• Analysis And Design Of Algorithms
• Computer Graphics

Military Commercial Joint Stock Bank (MBBank)

Data Scientist

April 2025 – Present | Hanoi

• Engineered a Positive-Unlabeled (PU) learning pipeline to overcome positive-only label constraints, achieving a 10x performance multiplier in high-value user acquisition compared to organic channels.
• Developed and optimized propensity models for targeted marketing, delivering a 200% performance uplift in user acquisition during rigorous backtesting.
• Orchestrated massive-scale data processing within a Data Lake architecture using PySpark and Polars.
• Architected and monitored a centralized Feature Store to streamline ML pipelines, utilizing Rust to ensure high-performance, low-latency Big Data processing and serving.
• Engineered cross-domain user personas (Term Deposits, Loans, Credit Cards) by synthesizing complex financial data, directly driving targeted business strategies.
• Deployed a robust fraud detection system leveraging anomaly detection algorithms to identify suspicious activities and mitigate financial risk.
• Awards: Honored with the "Honorable Staff Award" (H2 2025) for exceptional technical contributions and measurable business impact.

Grab Holdings Limited (GRAB)

Data Scientist (Full-time Intern)

May 2024 – Nov 2024 | Ho Chi Minh City

• Led end-to-end machine learning projects, from data preparation to production deployment, for user segmentation, recommender systems, and reseller identification.
• Collaborated with product and business teams to align model development with company goals, enhancing cross-functional impact.
• Achieved 100% growth in valuable user segments across food and logistics verticals by designing adaptable user persona models.
• Enhanced customer support efficiency with an RCA-based chatbot framework, achieving over 80% mapping accuracy using LLM-powered reranking and vector embeddings.
• Conducted A/B testing to validate model improvements and built scalable pipelines for continuous performance monitoring.

University of Information Technology, VNUHCM

Research Assistant

Nov 2023 – Jan 2025 | Ho Chi Minh City

• Authored "Towards Real-Time Open World Instance Segmentation", introducing a novel model with 23% recall for unknown instances, and proposed new benchmark.
• Developed MemoryFlow, a modular multi-agent LLM framework with vector database integration, enabling scalable and seamless LLM applications.

Programming Languages

Data Engineering & Analytics

Machine Learning & Deep Learning

MLOps & Cloud Platforms

Databases & Frameworks

LLM & AI Agent Frameworks