Machine Learning enthusiast currently learning and exploring NLP and Computer Vision fields. Committed to continuous learning and embracing challenges in the tech world.
Built backend APIs for AI applications and developed agentic workflows using MCP and orchestration frameworks. Fine-tuned open-source language models and contributed to AI research on LLMs and multi-agent systems.
Developed machine learning models and backend APIs for production-oriented applications. Designed and implemented multi-agent systems and ML solutions.
Presented "Learning Collaborative Reasoning Strategies Through Trust-Weighted Multi-Agent Consensus" at the WMAC Bridge Program on Advancing LLM-based Multi-Agent Collaboration.
Conducted research on improving efficiency and reducing the size of Small Language Models (SLMs) using scientific research methodologies.
Built a strong foundation in AI and gained hands-on experience through projects.
Learned data science fundamentals through Datacamp as part of the fellowship program.
End-to-end Nepali festive poster generator from English prompts using fine-tuned M2M-100 and Stable Diffusion v1.5, with custom text rendering and an automated generation pipeline.
YOLOv8 instance segmentation model for facial components including eyes, ears, nose, mouth, and eyebrows — designed as a foundation for facial similarity analysis.
Solves math word problems from text and image inputs using a fine-tuned BART model, OCR integration, and a recommendation module for similar practice problems.
Automated resume screening system that ranks candidates against job descriptions using FastText embeddings and weighted semantic similarity scoring.