-Collaborated on the development of basic to moderately complex SQL queries for campaign logic and reporting, optimizing data retrieval processes by 30%. -Utilized software tools, websites, and open-source research methods to identify relevant news coverage for government interests, improving news tracking efficiency by 25%. -Delivered high-quality daily news reports, Monday through Friday, maintaining a 95% on-time delivery rate. -Analyzed engagement data from News, Public Policy, and Healthcare publishers, identifying trends that led to a 15% increase in content interaction over six months. -Worked cross-functionally with Communications teams to share data insights with internal and external stakeholders, enhancing strategic alignment. -Troubleshoot data processes and values, partnering with IT to resolve system acceptance issues, achieving a 98% resolution rate on first-time troubleshooting efforts.

Research Works -Cross-Lingual Depression Identification Developed a multi-head attention network for explainable depression identification in Thai. -Psychological Well-Being Indicators Identified key indicators of psychological well-being through machine learning and explainable AI in Reddit interactions. -Thai Language Depression Detection Utilized attentive network models to detect depression in Thai posts, providing deeper mental health insights. -LGBT Cyberbullying Detection Applied transformer algorithms to identify cyberbullying in Thai content, specifically within the LGBT community. -SEO Transformation with Generative AI Explored the challenges and opportunities associated with transforming SEO in the age of generative AI.

Research Works : 1.Mutual Information-Based Logistic Regression for Phishing URL Detection: Developed a logistic regression model utilizing mutual information metrics to effectively identify phishing URLs, enhancing cybersecurity measures. Technologies Used: Python, scikit-learn, pandas, NumPy. 2.Thai-Language Chatbot Security: Implemented a detection system for instruction attacks in Thai-language chatbots using XLM-RoBERTa for text embeddings and Bi-GRU for sequence modeling, improving chatbot resilience against adversarial threats. Technologies Used: Python, XLM-RoBERTa, Bi-GRU, TensorFlow, Hugging Face Transformers, LLM Security. 3.Adversarial Learning for Mirai Botnet Detection: Leveraged Long Short-Term Memory (LSTM) networks combined with XGBoost to create an adversarial learning framework for detecting Mirai botnet activities, enhancing network security. Technologies Used: Python, LSTM, XGBoost, Keras, TensorFlow. 4.Speaker Recognition in Kannada Language: Adopted Mel-Frequency Cepstral Coefficients (MFCCs) integrated with a hybrid model of Random Forest and Multi-Layer Perceptron (MLP) for robust speaker recognition in the Kannada language, facilitating advancements in voice-based applications. Technologies Used: Python, MFCC, Random Forest, Multi-Layer Perceptron, scikit-learn, librosa.