Jay Patel

Ó - JAY PATEL - ¯ Jay Patel ‡ jayPatel029 Pune PROFILE Final-year Computer Engineering student and AI/ML enthusiast with hands-on experience in machine learning, deep learning, and Flutter app development. Skilled in building end-to-end AI solutions using TensorFlow, scikit-learn, and OpenCV, with a focus on real-world applications. Strong foundation in backend systems, mobile-first design, and MLOps basics. EDUCATION MIT Academy Of Engineering Jan 2022 – May 2025 Pune, Maharashtra Bachelor of Technology in Computer Engineering. (CGPA – 7.4) TECHNICAL SKILLS Languages: Python, Java, JavaScript ML & DL Frameworks: TensorFlow, Keras, scikit-learn, OpenCV Dev Tools & Frameworks: Flutter, Node.js, Git, Firebase, REST APIs Databases: MySQL, MongoDB INTERNSHIPS Full Stack Developer Intern Oct 2024 – Present Kifayti Health, Bangalore (Remote - Holistic care for chronic conditions) • Enhanced and maintained a healthcare app using Flutter and Node.js, ensuring scalability and performance. • Redesigned UI with Flutter optimizations, reducing load times and improving user experience. • Implemented JWT authentication and secured API communication, strengthening app security. • Integrated Firebase Cloud Messaging (FCM) and local notifications, enabling real-time alerts for better patient engagement. • Developed automated scheduled tasks for entry validation, ensuring data accuracy and reducing manual effort. MAJOR PROJECTS Multi-Modal Emotion Recognition System | (Completed) Deep Learning, TensorFlow, Keras LinkAug 2023 – Dec 2024 ∗ Built a real-time system using facial (FER) and speech (RAVDEES) data to detect user emotions. ∗ Used CNNs for facial expression recognition and LSTMs for speech-based emotion classification. ∗ Applied decision-level fusion to combine both modalities, boosting overall accuracy. ∗ Enabled live emotion tracking, suitable for mental health and customer service applications. QR Code Counterfeit Detection | (Completed) TensorFlow, Keras, OpenCV Link Feb 2025 – Mar 2025 ∗ Designed and trained a CNN model with MobileNetV2, achieving 94%+ accuracy in detecting counterfeit QR codes. ∗ Benchmarked performance against traditional methods using handcrafted features (HOG, LBP, ORB, Wavelets) with XGBoost, reaching 85% accuracy. ∗ Applied image augmentation and class balancing techniques to boost generalization. ∗ Optimized the model for deployment on mobile and edge devices, ensuring practical use in real-world scenarios. AI-Powered Onboarding Automation System | (Completed) Flutter, Django, OCR, Link Nov 2024 – Dec 2024 ∗ Built an AI-powered onboarding system to automate document processing and form filling using Tesseract OCR and Regex-based parsing. ∗ Reduced manual HR effort by improving form recognition accuracy and minimizing input errors. ∗ Developed a full-stack solution with a Flutter frontend and Django backend, using MongoDB for scalable data storage. MINI PROJECTS ∗ MCQForm App: Quiz application built with Flutter and HiveDB for offline MCQ storage. ∗ DocuVault: Secure document storage & retrieval web app using Flutter and Node.js. ∗ Corona Tracker App: Live COVID-19 tracking app leveraging REST APIs and Flutter. ∗ Care 32: Comprehensive dental care and appointment scheduling app using Flutter. ACHIEVEMENTS & CERTIFICATIONS AWS Academy Machine Learning Foundations Amazon ML Hackathon – Secured 397th place Developed ML pipeline for entity extraction from medical images (OCR + NLP)