Sarang Gade

Building web platforms, automation workflows, and ML-backed tools that solve real problems.

From a self-updating portfolio system to AI proctoring and maternal health ML, I like projects that combine product thinking with technical depth.

I am a Computer Science and Electronics student focused on building software that feels useful, thoughtful, and production-minded.

My work spans JavaScript, TypeScript, Python, Java, React, Next.js, Node.js, FastAPI, Flask, MongoDB, and machine learning, with a strong bias toward projects that do more than just look good in a demo.

I enjoy working on systems where automation, clean interfaces, and measurable outcomes matter.

const sarang = {
  currentlyBuilding: [
    "automation-first portfolio systems",
    "AI-assisted developer tools",
    "full-stack products with clean UX"
  ],
  learningMode: ["data structures", "system design", "ML deployment"],
  mindset: "Build it well, document it clearly, improve it continuously"
};

Languages JavaScript TypeScript Python Java

Frontend React Next.js HTML CSS

Backend Node.js Express.js FastAPI Flask

Data / ML scikit-learn SMOTE OpenCV MongoDB

Tools Git Docker GitHub Actions Vercel Render

Outcome: turned a personal portfolio into a self-maintaining system with automation, review workflows, and deployment hooks.

• Tech: Next.js Express.js MongoDB
• Build detail: added GitHub sync, Codolio screenshot automation, Discord notifications, manual fallbacks for LinkedIn data, and human-in-the-loop approval before production updates.
• Links: Repo · Live

Why it matters: this is more than a portfolio site; it is a productized workflow that reduces maintenance overhead and keeps personal branding continuously fresh.

Outcome: built an AI-based interview and exam proctoring system with live monitoring, alerting, and webcam-based behavior checks.

• Tech: Python FastAPI JavaScript
• Build detail: implemented camera selection, MJPEG streaming, CNN-based gaze checks, face-count detection, MediaPipe fallback logic, and a browser dashboard for live status and violations.
• Links: Repo

Why it matters: it tackles a hard real-world problem where accuracy, fallback behavior, and usability matter more than surface polish.

Outcome: built a full-stack ML application that predicts maternal health risk with 86.7% accuracy and 94.5% recall for high-risk cases.

• Tech: Python React Flask
• Build detail: compared 7 models, used SMOTE for class balancing, selected Gradient Boosting to reduce false negatives from 5 to 3, and wrapped the model in a React plus MongoDB workflow.
• Links: Repo · Live

Why it matters: it shows end-to-end thinking across ML, API design, frontend UX, and domain-specific evaluation where recall mattered more than raw accuracy.

• Logged 674 contributions in the last year on GitHub.
• Built across 27 public repositories covering frontend, backend, Java, automation, and machine learning.
• Earned GitHub profile achievements including Pair Extraordinaire, Pull Shark, Quickdraw, and YOLO.
• Built a deployed portfolio system at iamsarang.dev with automated update workflows.
• Shipped projects spanning AI proctoring, healthcare ML, portfolio automation, and DSA practice in Java.

• I prefer clear specs, fast iterations, and practical solutions over noise.
• Open to internships, student developer roles, and meaningful collaborations.
• Most interested in full-stack product engineering, automation systems, and applied machine learning.

If you're building something ambitious or looking for a developer who likes hard problems and fast learning curves, connect with me on LinkedIn or explore my work at iamsarang.dev.

_{Built for fast scanning, strong signal, and real project depth.}