Final-year Computer Engineering student at COEP Technological University. Currently interning at Pine Labs. Passionate about systems, full-stack web, and machine learning.
A passionate engineer who loves building things that matter
I'm Harshavardhan Prataprao Kale, a final-year Computer Engineering student at COEP Technological University, Pune. I thrive at the intersection of robust engineering and creative problem-solving.
From building a high-throughput C++ Database Engine to leading a full-stack team at a startup, I love challenging myself across domains — systems programming, web development, machine learning, and fintech infrastructure.
Currently contributing to Pine Labs' backend infrastructure using Java, while wrapping up my B.Tech with a strong focus on System Design and DSA.
My academic foundation and journey
COEP Technological University, Pune
Samarth Madhyamik va Uchhamadhyamik Vidyalay, Karjat
Dynamic English School, Karjat
Where I've worked and what I've built
Technologies, tools & frameworks I work with
A selection of things I've built — from systems to full-stack apps
RAG + Agentic AI system for intelligent loan decisions with sub-5s latency and RBI compliance-override safety. Multi-modal credit analysis.
SQL-compatible relational DB engine with ACID transactions, B-tree indexing, write-ahead logging, and MVCC. Achieved 3K QPS and 70% faster crash recovery.
Exclusive placement network for COEP students to share interview experiences, insights, and career guidance. Live on Vercel.
College social media platform with anonymous polling, "Connect to People" feature, interest-based recommendation algorithm, and WebSocket-powered real-time chat.
Multi-class sentiment classifier (positive/neutral/negative) achieving 85% accuracy using Logistic Regression on balanced Twitter data with TF-IDF/CountVectorizer.
Cross-platform expense manager with data visualization, automated budgeting suggestions powered by Gemini AI, and local-first Hive storage.
High-performance C++ file downloader using cURL and multithreading to download large files in concurrent chunks — significantly improved download speeds.
Full-stack attendance system with real-time face detection, role-based access for teachers/admins, and MongoDB backend for persistent records.
Open to opportunities, collaborations, and good conversations
kaleharshavardhan07@gmail.com
Pune, Maharashtra, India
Software Dev Intern @ Pine Labs