Saur Deshmukh

About Me

I am an enthusiastic Computer Engineering student (B.Tech, SPIT) with a strong foundation in full-stack development and machine learning. With hands-on experience in Python, React, Flask, Next.js, and AI model integration, I enjoy building impactful projects that merge software engineering with real-world applications. I thrive in collaborative hackathon environments and continuously push my boundaries through innovation and problem-solving.

Skills

• Programming Languages: C, C++, Python, JavaScript, Java
• Frontend Development: React, Next.js, Tailwind CSS
• Backend Development: Flask, Node.js, Express.js
• Databases: MySQL, MongoDB
• Tools & Frameworks: Git, GitHub, Docker, Postman

Achievements

• 🏆 Hackanova 4.0 (National Level Hackathon, TCET) – Domain Winner
• 🏆 SE Hack, SPIT – AI/ML Domain Winner
• 🎯 Barclays’ Hack-O-Hire 2025 – Finalist (Top 24)

Experience

Frontend Developer Intern @ SmowCode (June 2025 – July 2025)

• Built an STM32CubeIDE plugin using Java & SWT for cloud-based file sync and project configuration.
• Integrated Gitea with backend APIs using Apache HttpClient for project metadata, file tracking, and selective sync.

Projects

1. ThirdEye – Smart Surveillance for Safer Hostels

• Developed a multi-model AI surveillance system integrating ResNet, YOLOv8, DeepSORT, dlib for altercation detection, anomaly monitoring, and unauthorized entry.
• Built a parallelized real-time alert system with WhatsApp notifications and a monitoring dashboard.
• Ensured privacy compliance via gender detection and role-based access control.

2. FortiPhrase – Password Strength Analyzer & Passphrase Generator

• Engineered an offline password analyzer using a Decision Tree Regressor and local AI models (LLaMA3, Ollama) for secure password strength evaluation.
• Added PII detection and AI-generated passphrase suggestions to improve user security.
• Full-stack app with Next.js (Frontend) + Flask (Backend).

3. Signifique – AI-Powered Sign Language Translator

• Built a real-time Indian Sign Language translator (Flask + React) for sign-to-text & speech conversion.
• Created a custom ISL dataset using Mediapipe & trained a Random Forest Classifier (98% accuracy).
• Integrated AI-powered learning modules to make ISL more accessible.

4. VibeBoost – AI Mood Enhancer

• Developed an emotion-detection system with BERT to recommend memes, motivational quotes, and media.
• Integrated APIs (MemeGenerator, ZenQuotes) for dynamic content.
• Full-stack application using React + Django.