Kaviyathamizhan T K

Engineered and deployed production-ready ML services spanning NLP sentiment analysis, clinical diagnostic models, and time-series energy optimization pipelines.

• •Improved brand monitoring accuracy by fine-tuning BERT for multi-class sentiment classification and emotion detection on social media datasets, delivering structured JSON output via RESTful APIs.
• •Built a cardiovascular disease risk prediction model with feature engineering and class balancing; evaluated with F1-score and AUC-ROC for reliable diagnostic performance.
• •Implemented experiment tracking and model versioning with Weights & Biases, reducing model comparison time from hours to minutes.
• •Built an AI-driven energy optimization platform using LightGBM forecasting (RMSE 0.50, F1 0.786), engineering 16+ temporal features and implementing a SciPy HiGHS optimization engine for cost-efficient energy scheduling.

Accelerated enterprise logistics decision-making through data engineering and supply chain analytics.

• •Accelerated logistics decision-making by preprocessing enterprise datasets and building Tableau dashboards surfacing supply chain insights to cross-functional teams.
• •Performed automated exploratory data analysis (EDA) and data cleansing on multi-table industrial databases.
• •Collaborated directly with supply chain stakeholders to translate operational bottlenecks into measurable KPIs.

Completed intensive industrial technical training in Concepts of Computer Vision under the RAMP Scheme of the Ministry of MSME, Government of India.

• •Acquired practical mastery of core Computer Vision fundamentals, digital image processing pipelines, spatial filtering, and feature extraction algorithms.
• •Built and tested deep learning vision workflows for image classification, visual pattern analysis, and automated quality inspection.
• •Successfully completed comprehensive laboratory evaluations and practical assessments administered by NSIC Technical Services Centre.