Welcome to my GitHub! I'm Pakshal Bhandari, a passionate software engineer blending creativity and technical expertise in full-stack development and machine learning. With a Master's in Computer Science from Purdue University and a Bachelor's in Engineering from Visvesvaraya Technological University, Bengaluru, I thrive on solving complex problems with innovative solutions.
π― My Mission: To build impactful software and systems that simplify lives, optimize processes, and foster innovation.
- Programming Languages: Python π, R, Golang
- Domains: Full-Stack Development, Machine Learning, Image Processing, SQL
- Operating Systems: Linux π§, Windows, Mac
- Technologies: Django, Flask, React, MySQL, PostgreSQL, Docker, AWS, Azure, nltk, TensorFlow, sklearn, LangChain, Streamlit
- π 774F Consolidated data from 5+ databases using AWS Athena, creating Quicksight dashboards for actionable insights into customer demographics and purchase behavior.
- π οΈ Diagnosed and resolved 20+ critical SQL pipeline bugs, boosting data reliability by 30%.
- π€ Collaborated cross-functionally to ensure 99% report accuracy by resolving 50+ weekly data discrepancies.
- π§ Developed RESTful APIs with Django, reducing response times by 200ms and improving backend performance by 25%.
- β‘ Automated testing with Jenkins CI/CD pipelines, slashing deployment cycles by 35%.
- π Streamlined cryptographic infrastructure using HashiCorp Vault, enhancing security compliance and cutting setup times by 40%.
- π Automated data scraping and parsing, improving support team efficiency by 35% and halving issue resolution times.
- π Optimized data management processes with Python and HashiCorp Vault, reducing manual errors by 25%.
- π οΈ Created Python scripts for data scraping, parsing, and analysis, significantly boosting debugging efficiency.
- π€ Cold Email Generator: Personalized outreach email generation using LangChain, Python, Hugging Face Transformers, and Streamlit.
- π Prediction of Employee Attrition: Designed REST APIs in Django to predict employee turnover, leading to a full-time role promotion.
- π° Fake News Detection: Developed a multimodal solution combining Bi-LSTM, CNN, and Logistic Regression, achieving 95.8% accuracy.
- π£οΈ Pothole Detection & Geotagging: Built a YOLOv3 model to identify potholes and map coordinates.
- π§ LeetCode Helper: Automated LeetCode solution generator using Flask (backend), React (frontend), and Groq API for advanced problem-solving.
- π₯ Above and Beyond Award
- π Spot Award
- ποΈ Team Excellence Award
π© Have an idea or a challenge? Let's collaborate!
Feel free to:
- Email Me to discuss exciting opportunities.
- Connect on LinkedIn to network and share insights.