Sơn Nguyễn Văn Email & Phone Number

P. Nguyễn Trác, Yên Nghĩa, Hà Đông, Hà Nội

Responsible for teaching subjects related to data science and AI: NLP, Big data analysis, and artificial intelligence, Data visualization, ...Research the topics:1. Develop a small LLM able for mini-desktop2. LLM is a search and a combined information system for new knowledge.3. Meta-heuristic algorithm, DRL for combination optimization and timeseries data

B1, Haibatrung, Hanoi

- Research and build up the Large Language Model using TPUsOptimal and Approximation Algorithm:- Formulate the MIP, LP, and ILP for the optimization problems in Graph, WSNs.- Propose the approximate and approximation algorithm to solve the NP-Hard problem in the Graph and Graph application.- Guide a research group in building optimal models for NP-Hard Problems in WSNs.During this job, I have published 5 papers about the optimization algorithm and application in international journals and conferences.

35 Thi Sách Hanoi

I play a crucial role in researching, developing, and deploying NLP and site models for the EKYC and Customer Service chatbot. These models incorporate various modules, including Joint Learning of Intent and Named Entity Recognition (NER), Response Selection, Standardization of Vietnamese location names, and applying Large Language Models (LLM) for chatting functionality.Develop, research, and deploy the LLM and Large Multi-modal in the financial domain, along with building data systems for serving AI jobs:- Data collection and processing on C4, OSCAR- Multilingual large-scale dataset- Proposed Multilingual Pretrained Language Model for Financial Domain- Handle new LLMs in resource-limited environments: CPU apply and LLM indexThis work has contributed to a paper presented at the KSE conference, and a portion of the job has been transformed into a product within our internal chatbot system.Research and building model for Customer Profiling: Cross-Sale, Up-sale, Customer segmentation:- Develop a Clustering Model for user segmentation based on deep cluster- Develop a DFM model for a recommendation system for cross-sale and upsell products to current users.Research and build a model for time-series stock recommendation based on the expertise recommended.Technologies used: S3 Datalake, Python Backend, LLM training, RAG, Rasa chatbot, SQL, Docker, Streaming platform...

AI Engineer and Backend EngineerApplied Machine Learning and Deep Learning in Anomaly:- Research the anomaly in stream data- Apply Machine Learning and Deep Learning techniques: Autoencoder, Random Forest, and Isolation Forest to detect anomalies in streaming data.Technologies used: SQL database, Python Backend, Machine learning, ML ops, docker

Tokyo, Tokyo, Japan

Computer Network and Heuristic AlgorithmTopology Network:- Build the fault-tolerance models in Tori topology- Solve the fault-tolerance problem in miss-connection Tori by building pair-wise path

Backend EngineerTechnologies used: SQL database, Java Backend, HTML/CSS, docker

Master'S Degree, Computer Science, 3.9/4.0

Research about metaheuristic algorithms in solving the NP-Hard Problem Research the LLM model and serve it in the product by quantization.

Engineer'S Degree, Computer Science, 3.51/4.0