Elevated sentiment analysis accuracy by 10% via strategic dataset expansion and few-shot inference. Leveraged BiLSTM, advanced preprocessing, and word embedding

Developed a Language Translator with Hugging Face, achieving +0.8 BLEU through PEFT and LoRA fine-tuning. Expertise in model optimization and adaptation

Built a diabetes detection model using ML with preprocessing, SMOTE, and RandomForest. Achieved 100% accuracy, but caution advised for overfitting and data leak