Sandesh Reddy Kothi

Dynamic Sr. Data Scientist at DataFactz with expertise in AWS and advanced AI technologies. Achieved 90% accuracy in intelligent document processing and engineered real-time ML pipelines serving 500,000 predictions per hour. Proven ability to optimize data workflows and ensure compliance, demonstrating strong analytical and problem-solving skills.

AI Generated Text Detection:

● Engineered an NLP pipeline for AI-generated content classification using TF-IDF and Multinomial Naive Bayes, achieving 95.4% accuracy on unseen data.

● Applied advanced preprocessing techniques (tokenization, stemming, stop-word and accent removal) to enhance model performance.

● Developed and deployed an interactive, production-ready Streamlit interface for real-time AI-generated text detection.

● Aligned with AI trustworthiness trends by contributing to the detection of synthetic text across digital content platforms.

● Tools Used: Streamlit, Scikit-learn, NLTK, Pandas, NumPy.

Early Detection of Potato Diseases Using CNN:

● Designed and trained a convolutional neural network (CNN) to classify potato leaf diseases (Late Blight, Early Blight) with 96.4% accuracy.

● Deployed the model as a web application using Streamlit for real-time disease identification with average prediction latency < 5 seconds.

● Performed image augmentation (rotation, flipping) to improve generalization and robustness of the model.

● Promoted agricultural AI applications by facilitating precision farming and rapid disease response.

● Tools Used: Python, TensorFlow, Keras, Streamlit, NumPy, Pillow (PIL).