ML Engineer/Scientist
Education
- Ph.D in Computer Science, Kent State University, 2025
- M.S. in Computer Science, Kent State University, 2022
- B.E. in Computer Engineering, Kantipur Enginnering College, 2014
Research Experience & Professional
🧑🔬 Research Experience
Ph.D. – Kent State University
📍 Kent, OH | 2025
- Fesearch focused on **Applied AI on Mental Health doman **.
- Working on pretraining/finetuning LLM ** for **AI in Mental Health.
- Gradient Approximation based LoRA (Low Rank Adaptation Method).
Research Intern – Ph.D. 2022
📍 IBM Research, Yorktown Heights, NY | Summer 2022
- Developed a time-varying Causal Inference module (Counterfactual Simulator) for an open-source library.
- Published contributions to causallib on GitHub, contributing to causal inference research.
Research Intern – Ph.D. 2021
📍 IBM Research, Yorktown Heights, NY | Summer 2021
- Designed metrics to quantify research eminence in open-source communities.
- Built a BERT-based classifier to predict research impact using GitHub data.
🎯 Internship Experience
Full-Stack Developer – Research Project
📍 Kent State Public Health Department | Summer 2020
- Designed and developed a data warehouse & backend system for drug abuse prevention analytics.
- Integrated social media, healthcare, and census data for large-scale public health insights.
🛠 Professional Experience
Software Engineer
📍 Deerwalk Services Pvt. Ltd., Kathmandu, Nepal | June 2015 – July 2018
- Developed large-scale analytics tools for U.S. healthcare solutions.
- Implemented health insurance business logic using Hadoop jobs on Spark, deployed on AWS.
- Built high-performance web services using Elasticsearch for real-time healthcare data access.
- Created interactive dashboards with Groovy on Grails for healthcare insights.
Junior Big Data Software Engineer
📍 Tektak Services Pvt. Ltd., Kathmandu, Nepal | January 2014 – December 2014
- Developed the backbone of a large-scale messaging system using Hadoop, HBase, Elasticsearch, AKKA, and MQTT.
- Designed fault-tolerant distributed infrastructure for high-volume messaging.
Skills
- Programming & AI
- Python (PyTorch, TensorFlow, JAX)
- Foundational Models (Llama, Claude, GPT, T5, Mistral, BERT, ViT, CLIP, Terramind)
- Deep Learning (RNN/LSTM, CNN, Attention Mechanisms, Auto encoders, Constractive Learing, Active Learning, RL, RLHF)
- Pretraining & Finetuning (PEFT methods like LoRA)
- ML Libraries Huggingface, PyTorch, TensorFlow, Scikit-learn, JAX, pytorch-lightening, Transformers, Langchain, LlamaIndex, vLLM, Weights & Biases, Hyrda, Optuna, Deepspeed, CUDA, ROCm, HIP
- CUDA & Parallel Computing
- Big Data & ML Engineering
- *Distributed Computing (Hadoop, Spark, MapReduce, Databricks, Kafka)
- *ML Optimization (Model Compression, Quantization, Pruning)
- *Databases (MongoDB, HBase, ElasticSearch, Neo4j, SQL, Pinecone)
- Big Data & ML Engineering
- Git, Postman, MLFlow, Docker, Kubernetes, Copilot, Jira
- Jupyter, Anaconda, Linux
- Cloud Platforms (AWS, GCP, Azure)
Publications
Intersection of Big Five Personality Traits and Substance Use on X: Insight from the COVID-19 Pandemic
Maharjan J, Jin R, Zhu J, Kenne D, Intersection of Big Five Personality Traits and Substance Use on X: Insight from the COVID-19 Pandemic
Do Large Language Models (LLMs) Really Understand Personality? A Test of Embeddings vs. Zero-Shot (Preprint)
Maharjan J Do Large Language Models (LLMs) Really Understand Personality? A Test of Embeddings vs. Zero-Shot JMIR Preprints. 01/04/2025:75347
Differential Analysis of Age, Gender, Race, Sentiment, and Emotion in Substance Use Discourse on Twitter during the COVID-19 Pandemic: An NLP Approach
Maharjan J, Jin R, King J, Zhu J, Kenne D Differential Analysis of Age, Gender, Race, Sentiment, and Emotion in Substance Use Discourse on Twitter during the COVID-19 Pandemic: An NLP Approach JMIR Preprints. 08/10/2024:67333
Large-Scale Deep Learning–Enabled Infodemiological Analysis of Substance Use Patterns on Social Media: Insights From the COVID-19 Pandemic
Maharjan J, Zhu J, King J, Phan N, Kenne D, Jin R Large-Scale Deep Learning–Enabled Infodemiological Analysis of Substance Use Patterns on Social Media: Insights From the COVID-19 Pandemic JMIR Infodemiology 2025;5:e59076
Talks
Transformer-based LLMs
Conference proceedings talk at Kent State University, Kent OH, USA
Big Data and Hadoop Training
Talk at Deerwalk Institute of Technology, Kathmandu, Nepal
Teaching
Key Highlights
✅ Strong focus on Finetuning multi-modal Foundational Models
✅ Experience in end-to-end ML system design
✅ Industry + Research background (IBM + Ph.D. Research)
✅ Deep learning expertise in optimization, scalability, efficiency, and big data processing
✅ Publications in AI in NLP and vision, demonstrating research contributions