Summary
Certified AI Engineer with expertise in training, developing, and deploying deep learning models, including Large Language Models (LLMs). Proficient in supervised and unsupervised learning using Python, and experienced in building neural networks with Keras, PyTorch, and TensorFlow. Specialized in NLP applications leveraging frameworks like LangChain, Hugging Face, and generative pre-trained transformers (GPT), LLaMA, and BERT. Successfully implemented projects using Retrieval-Augmented Generation (RAG)
Work experience
Jokes Chat Bot Agent Feb 2025
Huggingface AI Agents Fundamentals Certificate
Built jokes chat bot agent that fetches jokes using API and tells jokes to user. It is also able to display time of specific country using tool that fetches the current local time in a specified timezone. Also it used latest Qwen/Qwen2.5-Coder-32B-Instruct as model. Text-to-image generation model was also used that can generate image from prompt. web_search, visit_webpage and final_answer are tools it is able to use, to perform actions required to provide response to user.
AI Tutor: Domain Expert in AGI / Formal Science
Mindrift | Freelancer April 2024 - November, Remote
As an AI Tutor Domain Expert in AGI / Formal Science at Mindrift, I develop specialized content that trains AI models to be ethical, accurate, and domain-specifically responsible. This pivotal work supports the reliability of AI applications in AGI /Formal Science. Responsibilities include crafting, editing, assessing, and fact-checking AI interactions based on rigorous research.
MetaMod2024 Workshop: Metabolic Modelling Sep 23 - Sep 26, 2024
In this workshop, we covered python based computation representation of metabolic network, flux balance analysis using linear programming. Programming based tool like Scrumpy was used along with Linux. Linear programming approach was used to investigate the core metabolism of Staphylococcus epidermidis as practical work to determine ATP production, biofilm formation and finally identify nutrients requirements for its growth. Similarly, for another practical, determination of minimal media requirements for Campylobacter jejuni was done. Here, we had done genome-scale metabolic model of Campylobacter jejuni M1cam to identify minimal media requirements and also understand the antimicrobial target identification. Moreover, identification of metabolic pathways involved in AMR in E.coli using GSMM was also done as practical work. GSM of E.coli was investigated. Experimental data was integrated into GSM and reactions, were identified. Hence
pathways involved in resistance against antibiotics was identified ultimately.
Fuse Machine Fellowship April 2024 - October 2024, Remote
As for fellowship participation, I would be getting hands on assignments and lectures related to Image Processing, Detection & Matching, CNN & Transfer Learning, Object Detection and Segmentation, Natural Language Processing, RNN & Transformers, Language Models & LLMs, Foundational Models , Structuring & Experiment & Deploying ML, Recommendation Systems, Prototyping with ML Services, MLOps, Reinforcement Learning, Time Series & Forecasting
Student | Tribhuvan University March 2023 - Present, Kathmandu
Machine Learning Assignments
- Implementation of multivariate linear regression from scratch
- Implementation of logistic regression to predict quantitative measure of progression on diabetes patient
- Implementation of Logistic Regression, SVM, Naive Bayes, KNN and Decision Tree for classification of quality of wine
- Implementation of stacking ensemble method
- Implementation of bagging method to train model for SMS spam detection
NLP Assignments
- Explore bag of words and frequency analysis to classify the news text
- Understanding n-gram, TF-IDF and document similarity using NLP methods
R Programming
- Social Network Analysis
- Application of different supervised and unsupervised machine learning algorithms
- Scraping tabular data from html and json to analyze using R
Monte Carlo Assignments
- Determination of area using acceptance-rejection method
- Random number generation using congruential method
- Determination of value of pi
- Integration of function
- Multi-variate integration
Education
Master of Data Science
Bioinformatics (Elective)
Bachelor of Civil Engineering
SKILLS
Python, R
SQL (MySQL), NoSQL (MongoDB)
Machine Learning, Deep Learning, Natural Language Processing (NLP)
EDA, Data Mining, Big Data, Statistical Analysis
Monte Carlo Simulations, Computational Statistics
Bio-informatics: Sequence Analysis, Phylogeny Analysis, Gene & Protein Prediction, Metabolic Modelling (Scrumpy)
GitHub, Linux
Tools
Mega11
AliView
FinchTV
FigTree
BLAST
ScrumPy
Languages
Nepali -Native
English -Fluent
Hindi -Intermediate