Open to collaborations & research
Dhiraj Pant
Computer Engineer
Kathmandu, Nepal
Computer Engineering graduate exploring how AI and technology can solve the real-world problems. Open for research and training collaborations where I can contribute, exchange knowledge and learn alongside like-minded individuals & teams.
About Me
B.E. Computer Engineering
2026 Graduate
Far Western University
Mahendranagar, Nepal
I'm a Computer Engineer (2026 graduate) with a deep interest in AI and applying it to real-world problems. Research gives me a way to explore ideas, test them, and understand the reasoning behind the results.
I enjoy building RAG pipelines, LLM applications, and deep learning systems. I've presented 2 research works at international conferences and earned NASA Space Apps 2025 Global Nominee recognition.
Outside of tech, I enjoy reading books and cricket.
What I Love
Want to know more about me, discuss an idea, or just have a chat?
Get In TouchWork Experience
AI Engineer @ AdVibe360 Pvt. Ltd.
1 Mar 2026 – 16 Sep 2026- Worked full-time on Generative AI solutions, developing LLM-powered applications and contextual chatbots integrated with backend applications, external APIs, and business workflows.
- Built end-to-end RAG pipelines covering document text extraction, chunking, embedding generation, vector indexing, and semantic retrieval to ground responses in domain-specific knowledge.
- Refined prompts, system instructions, and context structures to improve response relevance, consistency, and grounding.
- Used Python to develop AI application logic, data-processing components, and API integrations; tested and debugged retrieval quality, generated responses, and failure cases.
- Translated client requirements into AI workflows and prepared technical documentation, contributing across requirement analysis, workflow design, implementation, and testing.
- Served as Lead Instructor for a two-day AI Bootcamp at Balmiki Lincoln College, Birtamode, Jhapa, on 13–14 August 2026, teaching AI fundamentals, AI tools, Prompt Engineering, hands-on AI projects, and Responsible AI.
Skills & Expertise
Deep Learning
CNNs, Transformers
PyTorch
Model Development
TensorFlow
Deep Learning Framework
Keras
High-Level API
Scikit-learn
Classical ML
Transfer Learning
Fine-tuning & Pretrained Models
RAG Pipelines
LangChain, Retrieval
My Projects
NepAI News — Weekly AI News in Nepali
Newsletter delivering the week's biggest AI updates in simple Nepali every Monday at 9 AM. Aggregates and distills articles from 5+ trusted sources — OpenAI, DeepMind, Anthropic, Hugging Face, and more — so readers stop scrolling through dozens of articles.
Moodomic — AI Mental Health Platform
AI-based mental health management system integrating a conversational chatbot with CBT modules and personalized recommendations. Built a RAG system using LangChain for context-aware, accurate responses — designing and managing the knowledge retrieval pipeline end-to-end.
Vecrosoft — Plant Leaf Disease Detection
CNN-based deep learning model for real-time plant leaf disease detection. Designed a custom network architecture, optimized hyperparameters, and applied multiple regularization techniques to reduce overfitting — achieving a 3% reduction in error rate.
CO2 Emission & Temperature Anomaly Analysis
Cleaned, processed, and analyzed large-scale climate data to identify regional and global emission trends. Visualized findings through maps and charts in an interactive dashboard to support climate research — recognized as Global Nominee at NASA Space Apps Challenge 2025, Dhangadhi.
Exoplanet Classifier — ML App
Machine learning app that classifies exoplanet candidates from NASA Kepler mission data — distinguishing real planets from false positives. Built and deployed as an interactive Streamlit app.
Research Work
PlantNetV1: A Deep Learning-Based Approach for Plant Disease Detection
Presented at 2nd International Conference on Heritage, Innovation and Transformation
Proposed a custom CNN architecture for multi-class crop disease classification.
ResearchGate DOI 10.13140/RG.2.2.29179.43047DeepDefend: A Multi-Stage Approach for Suspicious Transaction Detection
Presented at 1st International Conference on Emerging Research Trends and Innovation
Designed a multi-stage deep learning pipeline for suspicious transaction and financial fraud detection.
ResearchGate DOI 10.13140/RG.2.2.30018.29120Latest Articles
Loading articles…
Awards & Certifications
Global Nominee — NASA Space Apps 2025
Dhangadhi · CO2 Emission & Temperature Anomaly Analysis
Analyzed large-scale climate data to surface regional and global emission trends in an interactive dashboard — recognized as a Global Nominee at the NASA International Space Apps Challenge 2025.
3rd Runner-up — Global IME AI/ML Hackathon
Global IME AI/ML Hackathon 2025
Placed 3rd runner-up at the Global IME AI/ML Hackathon 2025, competing against teams from across Nepal with an applied machine learning solution.
Machine Learning Specialization
DeepLearning.AI · Coursera
Supervised learning, neural networks, and practical ML project workflows — completed on Coursera.
Deep Learning Specialization
DeepLearning.AI · Coursera
CNNs, sequence models, and modern deep learning architectures — completed on Coursera.
Python for Everybody Specialization
University of Michigan · Coursera
Programming fundamentals, data structures, and web data access — completed on Coursera.
Contact Me
Send a Message
Get in Touch
Availability
Computer Engineer with an interest in AI. Open to research and training collaborations.