Ta Hai

Engineered robust Excel and PDF parsing pipelines for complex, domain-specific formats, optimizing Key Information Extraction (KIE) and downstream analysis. Architected advanced Text-to-SQL agents using LangChain and LangGraph, designing complex reasoning pipelines and workflows for accurate database query generation and validation. Engineered NLP pipelines for client projects, collaborating to execute advanced topic modeling and fine-tune domain-specific embeddings on real-time conversational transcripts.

A.I Engineer at Sun Asterisk (2026-02 – Present)

• Engineered robust Excel and PDF parsing pipelines for complex, domain-specific formats, optimizing Key Information Extraction (KIE) and downstream analysis.
• Architected advanced Text-to-SQL agents using LangChain and LangGraph, designing complex reasoning pipelines and workflows for accurate database query generation and validation.
• Engineered NLP pipelines for client projects, collaborating to execute advanced topic modeling and fine-tune domain-specific embeddings on real-time conversational transcripts.

A.I Engineer at Eureka Robotics (2025-07 – 2026-02)

• Reconstructed and optimized the data generation workflow to produce large-scale synthetic datasets for vision-based grasp detection in a data-scarce environment.
• Collaborated with the perception team to benchmark inference performance, reducing latency and memory footprint while maintaining model accuracy.

A.I Engineer at Viettel Telecom (2024-04 – 2025-05)

• Designed and implemented an end-to-end audio processing pipeline, including web scraping, preprocessing, and formatting raw audio into LJ-Speech dataset format.
• Conducted experiments and fine-tuning on various models for Vietnamese voice cloning, deploying solutions via Triton for real-time inference.
• Applied prompt engineering and text-chunking techniques to enable LLMs to summarize long dialogues into concise, high-quality outputs.

Bachelor of Science in Data Science and Artificial Intelligence – Hanoi University of Science and Technology (2026-08)