AI/ML Engineer & Researcher

Hi, I'm
Roshan Upadhyay

M.Tech student at IIT Patna specialising in AI & Data Science. I build intelligent systems — from computer vision and deep learning models to infrastructure automation and full-stack AI applications.

View Projects Get in Touch Blog
20+
Projects
9.44
CGPA (B.Tech)
2
IEEE Publications
3+
Years Experience

What I Know

Skills & Technologies

 Languages
PythonCC++ SQLHTMLCSS
 AI / ML
Machine LearningDeep Learning Reinforcement LearningComputer Vision NLPDSAOOP
 Frameworks & Libraries
TensorFlowKeras Scikit-LearnPyTorch OpenCVFlask NumPyPandas Matplotlib
 Tools & Platforms
LinuxPower BI ScriptingInfoblox ServiceNowF5 BIG-IP Google Earth EngineJupyter

Career

Work Experience

AI/ML & Infrastructure Engineer Aug 2024 – Present
 Microland Limited, Bangalore
  • Completed training across 12+ IT domains (Infrastructure, Networking, Cloud, DevOps, AI/ML, Security), strengthening cross-functional technical expertise.
  • Hands-on with Infoblox (IPAM/DNS/DHCP), ServiceNow, UltraDNS, F5 BIG-IP & Linux scripting; automated repetitive tasks, improving operational efficiency by 30%.
  • Performed IPAM data analysis on network address utilisation; identified unused/overlapping IP ranges, reducing IP wastage by 20% in lab environments.
  • Applied data-driven insights to infrastructure workflows, reducing manual effort by 25–40% through automation and structured data reporting.
ML Research Intern Jan 2023 – Mar 2023
 Indian Institute of Remote Sensing (IIRS), Dehradun
  • Developed predictive models using Python & ML algorithms, improving crop production forecasting accuracy by 20% in Punjab and Haryana.
  • Utilised Google Earth Engine to compute NDVI values and acreage for 50+ districts, enhancing agricultural planning efficiency by 15%.
  • Processed and analysed 10+ TB of satellite imagery, reducing data processing time by 25%.
Computer Vision Intern Sep 2022 – Feb 2023
 Theta Enerlytics Pvt. Ltd., New Delhi
  • Built a multi-object tracking model for video, enabling accurate detection across diverse scenarios.
  • Created a custom dataset to classify gunfire sounds using ML techniques for accurate categorisation.

What I've Built

Featured Projects

MEDIFORECAST
AI-based multiple disease prediction system using medical data and ML models, deployed as a Flask web app.
PythonFlaskScikit-LearnPandas
ICoach — Yoga Procoach
Personalised AI yoga coach using pose estimation with adaptive key point elimination for real-time feedback.
PythonOpenCVTensorFlowCNN
Driver Drowsiness Detection
Real-time fatigue detection using computer vision and facial landmark tracking to improve road safety.
PythonOpenCVKerasTensorFlow
Lunar Lander (RL)
Classic reinforcement learning simulation using OpenAI Gym's LunarLander environment to train an autonomous agent.
PythonOpenAI GymNumPyRL
Pneumonia Detection from X-Rays
Deep learning-based pneumonia detection from chest X-ray images using convolutional neural networks.
PythonKerasCNNOpenCV
Emotion Recognition
CNN-based deep learning model that detects human emotions from facial expressions in real time.
PythonOpenCVKerasCNN
Animal Detection (YOLO + R-CNN)
Object detection system identifying various animals using YOLO and Faster R-CNN models.
PyTorchYOLOFaster R-CNNOpenCV
SLAM Simulation
Simultaneous Localisation and Mapping where a virtual robot explores a floor plan and builds a real-time map.
PythonOpenCV
Image Caption Generator
Combines CNNs and RNNs to automatically generate descriptive captions for images using deep learning.
KerasTensorFlowCNNRNNNLP
Sentiment Analysis
NLP-based model performing sentiment classification (positive/negative/neutral) on text data.
PythonNLTKScikit-LearnNLP
Diabetic Retinopathy Detection
Deep learning system detecting diabetic retinopathy from retinal fundus images with Inception v4.
PythonKerasCNNOpenCV
Customer Lifetime Value Analysis
Predicts CLV using customer purchase history and ML regression — helps identify high-value customers.
PythonPandasScikit-LearnRegression

Academic Background

Education

Pursuing
M.Tech — AI & Data Science Engineering
Indian Institute of Technology, Patna
2026 – 2028
Completed
B.Tech — Artificial Intelligence
SRM Institute of Science & Technology
2020 – 2024
CGPA: 9.44 / 10

Research

Publications

Enhancing Diabetic Retinopathy Diagnosis with Inception v4: A Deep Learning Approach
 Published in IEEE
View on IEEE Xplore
Advanced Yoga Pose Estimation: Enhancing PoseNet with Adaptive Key Point Elimination
 Published in IEEE
View on IEEE Xplore
Patent: A System and Method for Real-Time Yoga Pose Detection Using a CNN Architecture
 Patent Application Publication
View Patent

Achievements

Awards & Certifications

 Awards
  • 3rd Place — MDP Exhibition, SRMIST 2024
  • POC Round — NASSCOM Academic Grand Challenge 2022
  • Final Round — Ideathon by CINTEL's NEXT-GEN AI 2022
  • 3rd Place — SRM Project Expo 2021
 Certifications
  • Machine Learning — Stanford University / Andrew Ng (Coursera)
  • IBM ML Professional Certificate — IBM on Coursera
  • Data Science & ML — CodingNinjas
  • Basics of Python — IIT Kharagpur (Enduro)
  • DSA in Python — CodingNinjas
  • Software Dev Processes — University of Minnesota (Coursera)
  • Intro to Python — CodingNinjas

Get In Touch

Contact Me

Let's Connect

Whether you want to discuss AI research, project collaboration, or just want to say hi — my inbox is open.