SUDHANTHIRAN
AI&ML Engineer, FOSS Enthusiast
AI&ML Engineer, FOSS Enthusiast
Hello, I'm Sudhanthiran N. I'm a AI enthusiast proficient in Computer Vision, LLMs and build many Machine Learning and Deep Learning solutions for real-time projects with pytorch and huggingface. Currenty I'm a freelancer working with IEEE research papers. I'm actively looking for AI related full-time roles and also happy in working with freelance projects.
I have worked on AI-powered content creation, job-seeking platforms, automated presentation generation, and PDF processing using LLaMA2, LangChain, PyTorch, and Hugging Face. My research internship at NIT Trichy has strengthened my expertise in cutting-edge AI technologies.
As the President of Odyssey Coding Club, I actively foster a tech community for AI enthusiasts. My goal is to leverage AI to create impactful solutions across various domains.
PowerPoint automation using LLaMA2, featuring PDF conversion and Flask interface. Creates precise, well-formatted presentations through AI-driven prompts, streamlining the design process.
One-shot platform parsing resumes and matching relevant jobs using PyPDF2, Langchain, and Huggingface Transformers. Features Django backend, Beautifulsoup webscraping, and Intel Devcloud deployment.
Project Details: Habitat Hive is a project that's aimed to solve rental finding problems. Habitat Hive uses a Habitability Index which ranks properties based on the living well-being around the property.
Project Details: This project enables a conversational AI interface for processing PDFs and answering user queries through retrieval-based techniques powered by large language models (LLMs).
Project Details: A CNNandLSTMbasedmodeldevelopedusingPytorchtogenerate descriptive prompts from images, inverting the Stable Diffusion Image generation approach..
Jun 2023 - Present
National Institute Of Technology Trichy, Tiruchirappalli
Trained object detection models like YOLOv8 and ResNet50 for fish activity classification, hitting 98.7% accuracy.
Curated, annotated, and preprocessed a dataset of 1000+ images, ensuring high-quality training data.
Fine-tuned LLMs (DistilBERT, Mistral-7B, Phi-2) on Reddit data to classify moral vs. immoral actions.
Explored multi-modal AI, integrating text and image-based models for better decision-making.
Worked with vector databases & retrieval-based models, optimizing AI-driven responses.
2021 - 2025
Saranathan College of Engineering
Successfully coordinated a workshop on GenAI with Google Cloud
Achieved Second place in Intel OneAPI Hackathon'23
Led the ODYSSEY coding club and oversaw the events
Trained Juniors for participating in various inter college Hackathons and events