Hi, I'm Sumeet Khatri
I'm a |
Final year B.Tech IT student and a Machine Learning Engineer with hands-on experience in building, deploying, and scaling end-to-end AI systems. Skilled in Computer Vision, MLOps, and developing Gen-AI solutions.
About Me
I am a passionate Machine Learning Engineer and a final year B.Tech Information Technology student at Indus University. I specialize in building and deploying end-to-end AI systems, with a strong focus on MLOps, Computer Vision, and real-time model serving using technologies like FastAPI and Streamlit.
My journey in AI has led me to explore the depths of Generative AI, including Langchain, Vector Databases, and building AI Agents. I have delivered impactful AI solutions, such as a credit scoring model achieving 92% AUC and a healthcare prediction system with 98% accuracy.
I am driven by the challenge of solving complex problems and am always eager to learn and apply new technologies to create efficient and scalable solutions.
Technical Skills
My Projects
LLM Fundamentals & Exploration
Explored tokenizer pipelines, dataset preparation, pretraining objectives, and transformer architectures like BERT & GPT.
News Summarizer (NLP)
Built an abstractive summarizer using BartForConditionalGeneration to generate concise, real-time news highlights. Deployed via Streamlit.
Toxic Comment Classifier
Multi-label NLP model using BERT to classify toxic comments. Deployed with a real-time Streamlit interface to flag abusive language.
Credit Risk Modeling
Engineered a credit risk model using LightGBM, achieving 92% AUC. Incorporated Weight of Evidence (WoE) and Optuna for tuning.
Car Damage Detection
Real-time damage assessment system using ResNet50 for multi-class image classification. Achieved 93% test accuracy.
Connect With Me
I'm currently looking for new opportunities. Feel free to reach out!