Hi π, I'm Vijay Shete
β’ Full-Stack AI Engineer and Researcher π§ About Me π¬ Research Intern at IIT Bombay: Focused on non-invasive eye-tracking technology for mental health assessment and depression screening. π« Healthcare AI Specialist: Building high-precision diagnostic engines for PCOS screening and multimodal medical assistants. π€ LLM & Multimodal Expert: Working with Llama-4, Whisper-large-v3, and Groq LPU for real-time speech and image reasoning. ποΈ Full-Stack ML Architect: Designing end-to-end systems from FastAPI backends to Dockerized cloud deployments.
πΌ Tech Stack
Category Tools & Technologies
π§ AI / ML, "Python, PyTorch, TensorFlow, Scikit-learn, CatBoost "
π£οΈ LLM & CV, "Llama-4, Whisper-large-v3, gTTS, CNNs, Multimodal LLMs "
π Data Science, "NumPy, Pandas, Matplotlib, Seaborn, SQL, Power BI "
π Systems, "FastAPI, Streamlit, Groq LPU (Inference Optimization) "
βοΈ MLOps & Tools, "Docker, Git/GitHub, Hugging Face Spaces, CI/CD "
π Featured Projects
ποΈ Multimodal AI Medical AssistantThe Tech: Meta-Llama-4, Whisper-large-v3, and Groq LPU API.The Impact: Analyzes X-rays/MRIs and provides real-time vocal diagnostic feedback with near-instantaneous processing.Deployment: Successfully deployed on Hugging Face.
π₯ Smart PCOS Screening SystemThe Tech: CatBoost (Diagnostic Engine), FastAPI, and Streamlit.The Results: Achieved 94% testing accuracy using clinical biomarkers with integrated NLP Chatbot support.Architecture: Containerized with Docker and automated via CI/CD pipelines.
π Publications & Research
Research Paper: "Industrial Revolution NLP to SQL Query Generator Using LLM" (IJIRCCE, 2025). Intellectual Property: Authored copyright for original research in NLP-to-SQL generation.
π« Connect With Me
πΌ LinkedIn: [https://www.linkedin.com/in/vijay442003/]
π§ Email: vijayshete442003@gmail.com
π Location: Pune, Maharashtra