NB.
Nomuundalai Batbileg

> Hello, I am

Nomuundalai
Batbileg

Data Scientist

Ulaanbaatar, Mongolia/XacBank

Building intelligent systems at the intersection of machine learning and software development. Specializing in production-ready AI solutions.

Get in Touch View Work

01. About Me

The Story So Far

I am a software engineer turned into a data scientist based in Ulaanbaatar, Mongolia. My journey began with frontend development, but my fascination with intelligent systems led me deep into natural language processing and computer vision.

At the Institute of Mathematics and Digital Technology, Mongolian Academy of Sciences, I spent over two years building and deploying machine learning models from neural machine translation to computer vision.

Today, I have been working as a Data Scientist at XacBank. I research and develop AI-driven solutions in financial sector.

4+
Years Experience
3
Deployed ML Models
4
Certifications
1
Published Scientific Paper

Education

Bachelor of Science in Software Engineering from Bauman Moscow State Technical University (class of 2022).

ML Expertise

2.2 years as ML Engineer at the Institute of Mathematics and Digital Technology, Mongolian Academy of Sciences.

Continuous Learning

Always researching the edge of AI/ML.

Science Background

Strong foundation in Mathematics and Physics, comfortable with core ML concepts.

02. Experience

Where I Have Worked

Data Scientist

Current

XacBank · Ulaanbaatar, Mongolia

2025.08 — Present

  • ▹I have conducted research on new technologies, designed an architectural diagram for AI agents and RAG.
  • ▹I developed an internal AI assistant for 1,200 employees and an external AI assistant for XacBank's main website. I integrated an internal AI assistant with the 'IThelpdesk' system, which is used daily by all IT employees.

Machine Learning Engineer

Institute of Mathematics and Digital Technology, Mongolian Academy of Sciences · Ulaanbaatar, Mongolia

2023.06 — 2025.08

  • ▹I developed models for text translation using the deep learning architectures such as 'seq2seq with attention' and 'transformer'. During model development, I prepared dataset for training, validation and testing.
  • ▹Accordingly, a copyrighted scientific article featuring the most accurate translation model has been published.
  • ▹I also fine-tuned the 'YOLOv10' models and performed object detection experiments.

03. Projects

Things I Have Built

External AI assistant

Built retrieval-augmented generation (RAG) workflow and developed an external AI assistant using LangFlow, Qdrant, LLM and Embedding model for XacBank’s main website.

    PythonPandasCrawleeLangFlowQdrantDockerGit (version control system)

    Internal AI assistant

    Built agent workflow and developed an internal AI assistant for XacBank’s 1200 employees. Integrated an internal AI assistant with our on-premises system “IThelpdesk” which is used daily by all IT employees.

      Copilot StudioPower AutomateWSO 2 API ManagerGit (version control system)

      Transformer-based translation model for two languages

      In this project, two types of architectures and three types of tokenization methods were used for the development between traditional Mongolian script and cyrillic Mongolian.

        PythonPandasNumpyTensorFlowPyTorchDeep Learning ArchitectureGit (version control system)

        04. Skills

        Tools & Technologies

        Languages

        Python80%
        SQL80%
        C / C++70%

        ML / AI

        PyTorch80%
        TensorFlow70%
        NumPy / Pandas90%
        Computer Vision60%
        Natural Language Processing70%

        DataBase & Tools

        PostgreSQL80%
        PowerBI90%
        Git / GitHub90%
        Docker70%

        05. Certifications

        Credentials & Learning

        Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization

        Coursera.org

        August 2025

        an online course authorized by DeepLearning.AI and offered through Coursera

        YEDM5QODEQT7 · Verified

        Neural Networks and Deep Learning

        Coursera.org

        July 2025

        an online course authorized by DeepLearning.AI and offered through Coursera

        DPFNXF76SR38 · Verified

        Artificial Intelligence and Machine Learning

        Mongolian University of Science and Technology

        May 2024

        Comprehensive AI/ML certification covering deep learning, computer vision, and NLP fundamentals

        ICTTC24-608 · Verified

        AWS Essentials and Cloud Computing

        Amazon Web Services

        April 2023

        Foundational AWS certification covering core cloud services, architecture, and best practices.

        XXX-XXX · Verified

        06. Contact

        Let's Work Together

        Whether you have a project in mind, want to collaborate on research, or just want to connect — I'd love to hear from you.

        Get in Touch

        Email
        nomuundalaibat@gmail.com
        Location
        Ulaanbaatar, Mongolia
        GitHub
        https://github.com/nomuundal-AI
        LinkedIn
        https://www.linkedin.com/in/nomuundalai-batbileg/