Hi there! 👋
I'm Abhay Narayan Dwivedi.
ML Engineer | Data Scientist | Web Developer. B.Tech student in Mathematics & Computing at Central University of Karnataka. Turning data into intelligent solutions.
ABOUT
I am a Mathematics & Computing undergraduate at the Central University of Karnataka with strong analytical foundations and hands-on experience in Reinforcement Learning, Deep Learning, Robotics Simulation, and Web Development.
I enjoy transforming raw data into actionable insights, building predictive models, and developing intelligent web applications. My work focuses on combining mathematical rigor with practical implementation to create efficient and scalable solutions.
Currently working on: Biped Locomotion using RL, AI-powered web applications, Predictive analytics systems.
My Journey
Journey Timeline
Intermediate
2023
Completed Intermediate at Kendriya Vidyalaya Mati, Kanpur Dehat.
Uttar Pradesh, India
B.Tech — Mathematics & Computing
2023 - Present
Pursuing B.Tech in Mathematics & Computing at Central University of Karnataka. Passionate about Reinforcement Learning, Deep Learning and building intelligent web applications.
Karnataka, India
Internship at IIT Mandi
May 2025 - Dec 2025
Developed and benchmarked a 6–8 DOF underactuated biped simulation using SAC, TD3, and DDPG with waypoint reward shaping (10M+ steps), achieving 2% fall rate and 94% navigation success.
Integrated A* global planning with SAC-based control for dynamic obstacle avoidance in simulation.
Himachal Pradesh, India
Published at ICC 2025 – Multi-Skill RL for Underactuated Biped
2025
Benchmarked SAC, TD3, DDPG using waypoint reward shaping (10M+ steps).
Achieved 2% fall rate, 94% navigation success in PyBullet simulation.
Research Project
SKILLS
Languages
Libraries
Frameworks
Tools & Simulators
Databases
PROJECTS
MNIST From Scratch
Built a fully connected neural network using only NumPy. Implemented forward/backward propagation and softmax manually. Achieved 90%+ accuracy on MNIST.
Customer Churn Prediction
ML models to predict customer churn with feature engineering. Achieved 78% ROC-AUC score.
Product Sales Analysis
Analyzed Amazon electronics sales data to uncover customer behavior patterns. Created insightful visualizations for business decisions.
TaskMate – Task Manager
Responsive task management web app with add/edit/delete tasks, dark mode, and real-time UI updates.
Hand Tracking Painter
Real-time gesture-based drawing app using OpenCV and MediaPipe. Computer vision for interactive user control.
Reinforcement Learning
Multi-skill locomotion in underactuated biped environments using RL. Focused on control systems and AI experimentation.