About Me

I am a Postdoctoral Fellow in the Department of Plant Pathology at Kansas State University, where I recently completed my Ph.D. in Computer Science. My current work focuses on building foundational Large Language Models (LLMs) for fungal genomes, using large-scale genomic sequence data and scalable HPC training pipelines.

My Ph.D. research focused on stance detection - developing advanced NLP models, especially transformer-based and Large Language Models (LLMs) - to analyze and classify opinions on contentious topics, with applications in social media analysis and public opinion tracking.

Download resume (PDF)

Education

Ph.D. in Computer Science

Kansas State University, Manhattan, KS

Dec 2025

M.S. in Computer Science

Kansas State University, Manhattan, KS

Aug 2025

Bachelor’s Degree in Computer Engineering

Tribhuvan University, Nepal

Oct 2016

Experience

Postdoctoral Fellow

Department of Plant Pathology, Kansas State University, Manhattan, KS

Sep 2025 - Present

  • Pre-training a foundational LLM model with up to 1.3B parameters for fungal genomes using large-scale genomic sequence data.
  • Building scalable HPC/ACCESS training pipelines for genome-scale data processing and model optimization.
  • Collaborating with computer scientists and plant pathologists to integrate biological insights from model development.

Graduate Research Assistant

Machine Learning & Data Science (MLDS) Lab, Kansas State University, Manhattan, KS

Jan 2023 - Aug 2025

  • Developed advanced transformer and LLM models for NLP tasks like stance detection across multiple domains, including tweets and financial text data from quarterly call reports.
  • Leveraged Retrieval-Augmented Generation (RAG) and prompt engineering techniques to extract valuable insights from Large Language Models (LLMs).
  • Optimized LLMs using Reinforcement Learning from Human Feedback (RLHF) to enhance response quality and alignment.
  • Implemented state-of-the-art AI research from paper to code, adapting it for custom research requirements.
  • Built and optimized large-scale machine learning models using Python and modern ML frameworks, using LLMs to improve semi-supervised baseline accuracy by 10% on average on a challenging dataset.
  • Collaborated on multiple ML/NLP projects and published research papers in top-tier conferences.

Graduate Teaching Assistant

Department of Computer Science, Kansas State University, Manhattan, KS

Aug 2019 - Dec 2022

  • Created quizzes, graded assignments, and provided academic support to students on AI, machine learning, and deep learning courses and term projects.
  • Developed and evaluated technical assignments and mentored students on programming and database projects for Programming Languages, Database System, and Computer and Information Security courses.

University Research Internship

LiuLab, Department of Plant Pathology, Kansas State University, Manhattan, KS

May 2021 - Aug 2021, May 2022 - Aug 2022

  • Applied Recurrent Neural Networks (RNNs), specifically Bi-LSTM models, to classify DNA sequences as core- or mini-chromosomes in Magnaporthe oryzae strains.
  • Developed and optimized a deep learning pipeline for genomic sequence classification, achieving high accuracy in predicting mini-chromosome presence.
  • Collaborated with an interdisciplinary team of computer scientists and plant pathologists to publish findings in a peer-reviewed journal.

Security Analytics Engineer

Logpoint, Kathmandu, Nepal

Aug 2017 - Jul 2019

  • Developed machine learning models for User and Entity Behavior Analytics (UEBA).
  • Developed algorithms for security log analysis and pattern detection, and maintained production-grade software systems.
  • Developed, maintained, and optimized plugins and code to efficiently parse thousands of logs using Python Regex, improving log parsing speed by 10%.
  • Took complete ownership of the development lifecycle for the product developed.

Software Developer Intern

Leapfrog Technology, Kathmandu, Nepal

Dec 2016 - Feb 2017

  • Created a Video to Gif converter using machine learning and neural networks.
  • Created various website templates using HTML5, CSS, and JavaScript.
  • Collaborated with back-end developers, UI/UX designers, and managers in a professional, corporate environment to meet client requirements.

Publications

  1. Jha, Aditya Nath, Markus Hoehn, Andrei Mazin, Nikesh Gyawali, H M Abdul Aziz, Doina Caragea, et al.
    International Conference on Transportation and Development 2026, pp. 60-71, 2026
  2. Gyawali, Nikesh, Doina Caragea, Cornelia Caragea, and Saif M. Mohammad
    PLOS ONE 20 (12): e0337911, Dec 2025
  3. Gyawali, Nikesh, Doina Caragea, Alex Vasenkov, and Cornelia Caragea
    arXiv preprint arXiv:2510.23464, Oct 2025
  4. Gyawali, Nikesh, Sarthak Khanal, Doina Caragea, H M Abdul Aziz, and Eric J. Fitzsimmons
    Journal of Transportation Safety & Security 17 (12): 1580-1611, 2025
  5. Gyawali, Nikesh, Iustin Sirbu, Tiberiu Sosea, Sarthak Khanal, Doina Caragea, Traian Rebedea, and Cornelia Caragea
    Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 12027-12044, Aug 2024
  6. Gyawali, Nikesh, Yangfan Hao, Guifang Lin, Jun Huang, Ravi Bika, Lidia Calderon Daza, Huakun Zheng, et al.
    NAR Genomics and Bioinformatics 6 (3): lqae108, 2024
  7. Taghian Dinani, Soudabeh, Doina Caragea, and Nikesh Gyawali
    Communications in Computer and Information Science 2105 (DATA 2023), pp. 73-94, 2024
  8. Tasali, Qais, Nikesh Gyawali, and Eugene Y. Vasserman
    Proceedings of the 7th Symposium on Hot Topics in the Science of Security, Sep 2020

Technical Skills

Programming Languages

PythonBashJavaScriptSQL

Machine Learning / Data Frameworks

PyTorchTensorFlowTransformersScikit-learnApache SparkHugging Face

Specializations

Machine / Deep LearningNatural Language Processing (NLP)Large Language Models (LLMs)

Data Visualization

MatplotlibSeabornPlotlyPowerBI

Tools & Technologies

CUDARegexDjangoHTML5CSSBootstrapHigh-Performance Computing (HPC)

Others

AWS / Cloud PlatformETL PipelinePrompt EngineeringRAGRLHFLangChainOllama