Headline

AI & Machine Learning Fresher | Built Data-Driven Models on 10k+ Records | Generative AI, NLP, Predictive ML | Data Analytics & Visualization

About

AI & Machine Learning fresher with a strong interest in Artificial Intelligence, Machine Learning, Data Analytics, and Data Science, focused on building data-driven, scalable, and reliable AI solutions. I have hands-on experience across predictive analytics, NLP, LLMs, deep learning, and applied ML systems, with a results-oriented approach to problem-solving. My core interests include AI chatbots, NLP-based systems, predictive ML models, CNN-driven analytics, and data-driven decision support, where I convert raw data into accurate models, actionable insights, and deployable applications. TECHNICAL SKILLS: Machine Learning, Predictive Modeling, Generative AI, NLP, CNNs, Feature Engineering, EDA, Model Evaluation, Data Visualization | RAG, Hugging Face, Vector Databases (Pinecone) | Python, NumPy, Pandas, Scikit-learn, TensorFlow, PyTorch, SQL | Streamlit-based ML applications | Hands-on experience with datasets ranging from 500 to 10,000+ records/images, focusing on accuracy, reliability, and production awareness. PROJECTS • MediBot – AI Health Companion: Built an NLP-based intent classification system across 20+ medical domains with multilingual support and safety validation to improve symptom triage. Achieved 92% accuracy, reduced misclassification by 30%, and handled 1,000+ real-world queries. • CNN-Based Medical Image Analytics (MRI): Developed an unsupervised CNN autoencoder on 10,000+ MRI images, integrating confidence scoring and drift simulation to address unreliable predictions. Improved model reliability under noisy and unseen data. • Airline Fare Predictive Analytics: Engineered Random Forest and XGBoost models on 10,000+ flight records with temporal demand features. Achieved R² > 0.91, improved accuracy by 28%, and enabled sub-second predictions. • TaxMate – AI-Powered Tax Assistant: Automated tax query resolution using intent classification and workflow automation. Reached 95% intent precision, automated 50+ scenarios, and reduced manual lookup time by 70%. INTERNSHIPS [IBM SkillsBuild (AICTE)] • AI/ML Intern: Built and deployed a salary prediction model (85% accuracy) on 500+ records, delivering 30% faster predictions via a Streamlit app. • Front-End Intern: Developed an educational platform with 50+ modules, improving load speed by 20% and maintainability by 40%. WORKSHOPS • AI Prompt Engineering (rapid AI prototyping) • AWS Cloud Hosting & Load Balancing (fundamentals) CAREER OBJECTIVE: Seeking entry-level roles in ML, AI, Data Analytics, or Data Science to contribute to data-driven innovation and grow as an industry professional.

Location

Chennai

Work experience

Artificial intelligence and Machine Learning Intern at IBM, Jun 2025 - Jul 2025 · 2 mos, Built an end-to-end Machine Learning–driven income classification system to predict whether an employee’s annual salary falls into ≤50K or >50K categories using real-world workforce data. The project addressed key challenges such as manual salary assessment, data inconsistency, bias, and lack of scalability in traditional decision-making processes. Performed extensive data preprocessing on a large structured dataset, including missing value handling, categorical data normalization, redundancy removal, and outlier elimination, reducing noise and improving data quality by ~20–25%. Applied Label Encoding for multiple categorical attributes and Min-Max Scaling to ensure feature uniformity across models. Designed and evaluated multiple supervised learning models — KNN, Logistic Regression, Multi-Layer Perceptron (Neural Network), and Random Forest Classifier — and conducted comparative performance analysis using accuracy score, confusion matrix, and classification report. The Random Forest model outperformed others, achieving ~85–88% prediction accuracy, improved generalization, and lower misclassification rates due to its ensemble-based learning approach. Implemented a real-time, user-input–based prediction workflow, allowing dynamic salary group prediction based on features such as age, education level, occupation, workclass, weekly working hours, and capital gain/loss, making the solution practical and decision-ready. Optimized the solution for deployment readiness by serializing the best-performing model using Joblib, enabling seamless integration into web applications, HR analytics tools, or enterprise AI systems. This project demonstrates strong hands-on expertise in Machine Learning pipelines, data-driven modeling, feature engineering, model evaluation, and production-oriented AI solution development, aligning with industry-grade ML workflows.

Data Science Intern at Bharat Intern, Jun 2024 - Jul 2024 · 2 mos, Remote

Internshala Student Partner at Internshala, Jun 2024 - Jul 2024 · 2 mos, Remote, Selected as an Internshala Student Partner, leading campus initiatives to promote internships, skill development, and career opportunities among students. Successfully engaged 500+ peers, organized awareness campaigns, and facilitated access to 100+ internship opportunities, improving student participation and placement readiness. Gained hands-on experience in digital outreach, student engagement, marketing strategy, and event coordination, while building strong communication, leadership, and project management skills.

Front-End Development Intern at IBM, Jun 2024 - Jul 2024 · 2 mos, Developed StudySlexia, a full‑featured interactive learning web platform inspired by GeeksforGeeks, aimed at enhancing programming education and technical skill acquisition. The system integrates multi‑category course access, structured learning paths, and user authentication to facilitate scalable self‑paced learning. Implemented responsive UI/UX design, efficient frontend routing, and dynamic content presentation to improve engagement across desktop and mobile devices. Built using HTML, CSS, JavaScript, and modern web development best practices, this project showcases expertise in frontend design, DOM manipulation, responsive layouts, and web interactivity. Successfully structured a multi‑section learning experience covering topics such as programming, interview prep, and technical categories, enabling users to explore and interact with content intuitively. Demonstrated the ability to replicate core features of a real‑world education portal, optimize component reuse, and enhance overall usability, resulting in an accessible learning resource that mirrors leading tech education sites. This project highlights strong capabilities in web development, frontend architecture, interactive UI design, and project delivery aligned with industry standards.

Education

Bachelor of Technology - BTech, Information Technology, BSA Crescent Institute of Science and Technology

kendriya vidyalaya