Likitha G

Led the development of large language model-powered chat experiences and AI-driven fraud detection systems to improve customer service and financial security. Integrated advanced frameworks like LangChain and Llama Index for retrieval-augmented generation pipelines, enhancing model accuracy. Developed AI-driven credit risk and financial coaching systems utilizing transformer architectures. Managed MLOps workflows for model deployment and ensured adherence to explainability, privacy, and fairness standards. Supported backend integration using Kotlin and collaborated with legal and risk teams for compliance.

Built LLM-powered chat experiences using the Kore.ai platform, enhancing self-service across credit card servicing. Integrated Retrieval-Augmented Generation (RAG) pipelines to improve model accuracy and transparency. Developed generative AI-powered fraud detection to simulate new fraud patterns and reduce false positives. Designed AI credit risk assessment models using transformer architectures for nuanced decisions. Automated financial coaching systems generating personalized advice. Pioneered Explainable AI to ensure transparency for customers and regulators. Managed MLOps workflows and collaborated with risk/legal teams to ensure compliance and privacy across AI solutions.

Developed deep learning pipelines for medical video and image analysis supporting early detection of respiratory and cardiac conditions. Built intelligent voice/chat assistants and retrieval-augmented generation systems for clinical protocols. Integrated Azure Cognitive Services for processing unstructured health records. Developed AI-powered ultrasound and MRI analysis tools, predictive analytics for patient monitoring, and NLP systems for clinical documentation. Maintained ML infrastructure, collaborated on responsible AI practices compliant with healthcare regulations, and developed NLP proof-of-concept models on AWS.