Data Science with less than a year in Machine Learning & AI
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Data Scientist and Machine Learning enthusiast with hands-on experience building end-to-end AI applications and data-driven solutions. Developed projects including a face recognition attendance system, customer churn prediction platform with retention recommendation logic, and interactive Streamlit dashboards integrated with APIs. Skilled in Python, Machine Learning, Data Analysis, and model deployment, with strong analytical and problem-solving abilities. Passionate about leveraging AI and data to solve real-world business challenges and building impactful, scalable solutions.
Dibrugarh University
Integrated M.Sc. · Physics - Space & Atmospheric Science
January 1, 2019 – January 1, 2024
Salt Brook Academy
High School (Class X) · HSLC
N/A – January 1, 2017
Salt Brook Academy
Higher Secondary (Class XII) · Science
N/A – January 1, 2019
Customer Segmentation & Clustering System for E-Commerce Personalsed Marketing
June 18, 2026 – Present
Analysed e-commerce transactional data to segment customers using unsupervised learning for data-driven marketing decisions. Applied RFM (Recency, Frequency, Monetary) analysis to engineer meaningful behavioural features from raw purchase history. Implemented K-Means and DBSCAN clustering algorithms; used Elbow Method and Silhouette Score to determine optimal cluster count. Derived actionable customer personas (e.g., high-value loyal customers, at-risk churners, new buyers) to inform targeted retention campaigns and personalised recommendations.
Deep Reinforcement Learning Agent for Flappy Bird
June 18, 2026 – Present
Developed a Deep Q-Network (DQN) based reinforcement learning agent to autonomously play the Flappy Bird environment using the Gymnasium framework. Implemented experience replay memory and target network synchronization to stabilize Q-learning training. Designed epsilon-greedy exploration strategy with configurable decay scheduling for optimized exploration-exploitation balance. Built a modular RL training pipeline with GPU/MPS acceleration support using PyTorch. Engineered hyperparameter-driven experimentation framework using YAML configuration files. Integrated model checkpointing, reward tracking, and logging system for training evaluation and reproducibility. Optimized neural network training workflow with mini-batch learning and Bellman equation-based Q-value updates.
SnapClass – voice and face recognition attendance system
June 18, 2026 – Present
Developed an AI-powered attendance management system using Face Recognition and Voice Recognition for automated student verification. Implemented real-time facial detection and voice-based authentication to improve attendance accuracy and reduce proxy entries. Built an interactive user interface using Streamlit for seamless attendance monitoring and management. Integrated Python-based machine learning and computer vision libraries for biometric processing and recognition. Designed a scalable system capable of storing and managing attendance records efficiently for educational environments.
Loan Approval Prediction System
June 18, 2026 – Present
Built a binary classification pipeline to predict loan approval status based on applicant financial and demographic features. Performed exploratory data analysis (EDA), handled missing values, encoded categorical variables, and applied feature scaling for model readiness. Trained and evaluated multiple classifiers including Logistic Regression, Decision Tree, Random Forest, and XGBoost; selected best model using ROC-AUC and F1-score. Addressed class imbalance using SMOTE oversampling to improve recall for high-risk loan cases, reducing false negatives.
Comparison of Machine Learning Regression Models for Temperature Prediction
June 18, 2026 – Present
Collected and pre-processed historical atmospheric temperature datasets for model training and evaluation. Implemented and benchmarked multiple regression algorithms including Linear Regression, Random Forest, Gradient Boosting, and SVR. Performed hyperparameter tuning and cross-validation to optimise model performance metrics (RMSE, R²). Visualised model comparison results to derive data-driven insights for climate prediction tasks.
Artificial Intelligence & Machine Learning
ApnaSchool
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's project portfolio shows a strong interest in applying AI/ML to various domains, from climate prediction to e-commerce and biometric systems. This diversity indicates a curious and adaptable mindset. The academic background in Physics, combined with practical ML projects, suggests a blend of theoretical understanding and practical application. The professional projects align well with a Data Science role, demonstrating an ability to tackle real-world problems. However, the lack of traditional work experience (experienceLevel: 0) means cultural fit related to a corporate environment, team dynamics, and long-term project contributions is largely unproven.
Soft Skills & Operational Fit
The candidate demonstrates strong analytical and problem-solving abilities, as evidenced by their project descriptions and academic background. Their experience with diverse project types suggests adaptability and a proactive approach to learning new technologies. The detailed project descriptions indicate good communication of technical concepts. However, without direct interview data or psychometric test results, it's difficult to fully assess stress handling, team collaboration, or specific work attitudes.