About
Computer Science (AI & ML) undergraduate with hands-on experience building end-to-end machine learning and full-stack applications across clustering, recommendation systems, and classification. Skilled in Python, Java, SQL, scikit-learn, Flask, and JavaScript, with current industry experience as a Machine Learning Intern at Naviotech.
Experience
-
Machine Learning InternNaviotechAug 2002 – Present
Completed a one-month hands-on Machine Learning using Python training program covering data preprocessing, model development, evaluation, and implementation. Developing an assigned end-to-end property-intelligence platform using K-Means clustering, Flask, scikit-learn, and interactive visualizations.
Education
-
B.Tech. in Computer Science (Artificial Intelligence and Machine Learning)CMR Institute of Technology, Hyderabad · Jawaharlal Nehru Technological University, Hyderabad (JNTUH)Computer Science, Artificial Intelligence and Machine Learning · 2023 – 2027
-
12th (Intermediate)Excellencia Junior College, ShamirpetScience · 1990 – 2023
-
10th (SSC)Jain Heritage a Cambridge School, Majeedpur2020 – 1996
Skills
Tools / apps / platforms
Projects
-
Python, Django, scikit-learn, Pandas, CountVectorizer, MySQL
Built a Django-based fraud-detection application using a 1,416-record credit-card transaction dataset, integrating model training, prediction, and result management. Transformed transaction identifiers with CountVectorizer and used a reproducible 67/33 train-test split for supervised binary classification. Implemented and compared Naive Bayes, SVM, Random Forest, and Gradient Boosting classifiers using accuracy, confusion matrices, and classification reports. Built a hard-voting ensemble combining Naive Bayes, SVM, and Random Forest to classify new transactions as fraudulent or non-fraudulent through the web interface.
-
Python, scikit-learn, deep-learning framework, API backend, database, D3.js/Plotly
Designing a context-aware recommendation engine for visual analytics dashboards that adapts suggestions using user session behavior and task context. Addressing cold-start scenarios by incorporating behavioral signals such as clicks, dwell time, navigation patterns, and query interactions alongside content signals. Developing embedding- and attention-based multimodal fusion to combine behavioral and content signals into a unified user-item representation. Planning evaluation against Collaborative Filtering and Content-Based Filtering baselines using nDCG, Precision@K, and Recall@K across warm-start and cold-start scenarios. Building the recommendation pipeline and visual analytics interface as an ongoing team project.
-
Python, Flask, scikit-learn, Pandas, NumPy, HTML/CSS/JavaScript, Plotly.js, Leaflet.js
Built a full-stack property-intelligence web application that segments 12,847 cleaned rental listings across Delhi, Mumbai, and Pune into distinct market communities. Engineered a 5-dimensional similarity space using log1p variance stabilization and StandardScaler; selected K=5 through Elbow and Silhouette analysis across K=2–10 and profiled the resulting segments with dynamically computed "Segment DNA" fingerprints. Designed a zero-data-leakage property-matching engine and NearestNeighbors-based similar-property finder by transforming new inputs through the fitted preprocessing pipeline without refitting. Shipped 6 interactive pages backed by a REST API with Leaflet.js maps and Plotly.js charts, and added automated tests for data integrity, clustering correctness, geospatial bounds, and matching logic.
Courses & certifications
- Machine Learning Training Completion · Naviotech Solution Pvt Ltd · 2026
- Claude 101 · Anthropic
- Retrieval-Augmented Generation for Enhanced AI Outputs · IBM SkillsBuild
- Make Agentic AI Work For You · IBM SkillsBuild
⚽ Extracurricular activities
-
IKSHANA Social Service Club Member
Active member involved in donation drives, volunteering, event management, and college technical workshops.