About Me
Go thru this if you don't wanna scroll all way down...😉
Learning from Errors!
Hey there, I'm Suyog Khanal, your friendly neighborhood Data Scientist! I hold a Master’s in Data Science from Vellore Institute of Technology (GPA: 9.6/10) and a Bachelor’s in Data Science & Analytics from Sharda University (GPA: 9.71/10, Dean’s List).
Currently, I’m working as a Data Scientist at Flodata Analytics, where I’ve engineered YOLOv10 models for object detection, built a Vision Transformer (ViT) for exercise recognition, and developed an offline AI meeting assistant using OpenAI Whisper, Vosk, and Google Cloud Speech-to-Text. I’ve also optimized Meta’s SAM 2 for image segmentation and designed an OCR pipeline using GPT-4o vision for handwritten recognition.
Before this, I was an AI Engineer Intern at Techolution, where I integrated AI functionalities in an online IDE, deployed a Retrieval-Augmented Generation (RAG) Q&A system, and built a cross-language code conversion feature with Anthropic Claude API. Earlier, as a Machine Learning Engineer Intern at KUKU FM, I designed a FastAPI-based Recommendation System, implemented Marketing Mix Modeling (MMM) using Meta’s Robyn, and developed an audio embedding system with OpenAI Whisper.
My Master’s thesis, BotanicaSpeak, focused on designing an intelligent LLM-powered chatbot for agriculture and forestry using LLaMA 2, RAG, and AWS SageMaker. Alongside, I’ve built impactful projects like a Cataract Detection App (99.3% accuracy with MobileNetV2), a Cricket Team & Player Recommender, and a Fruit Freshness Classifier achieving 98% accuracy.
On the research front, I’ve published a paper on Deep Learning for Fruit Classification (Indian Journal of Natural Sciences, 2023) and authored a study on EEG-based Sleep Apnea Detection that’s under review in the Journal of Artificial Intelligence in Medicine.
My technical toolkit includes Python, R, TensorFlow, NLP, Computer Vision, SQL, Docker, Azure, AWS SageMaker, and Large Language Models. Whether it’s designing cutting-edge AI systems or solving real-world data challenges, I’m passionate about crafting futuristic, impactful solutions.
