This is Santhosh Reddy BasiReddy, around 10 years of experience as Senior Salesforce Architect and a full-stack engineer with strong background in designing enterprise application, modern web/mobile frameworks and distributed systems. My work in recent times focusses on Natural language programming (NLP), Large Language Models (LLMs) and Geni Ai Engineering exploring how AI Architecture can be integrated to real world mobile, cloud and enterprise applications. My Professional experience includes designing and scaling large enterprise systems, most recent project Risk control and self-Assessment (RCSA) at USAA, where I have designed, architected and implemented a high-performance audit ready compliance solutions using salesforce, JavaScript and modern UI frameworks. This experience made me investigate on the areas where Gen AI can be implemented in RCSA Tool within regulated areas, I bring together my business experience with hands on engineering skills across Java, JavaScript/TypeScript, Angular, React, React Native, Node.js, and Firebase to design solutions that seamlessly bridge traditional software engineering with cutting-edge AI/LLM capabilities. GenAI for RCSA (Risk and Control Self-Assessment) I am currently researching how GenAI can enhance key components of the RCSA process by generating intelligent, context-aware suggestions that help risk owners work more efficiently and consistently. My focus includes exploring how LLMs can assist with: • Using historical RCSA data, learning patterns and Suggesting ratings, descriptions while doing assessment. • Risk identification using AI assisted from process narratives • Risk and control taxonomy mapping across business units and regions • Recommendations on control design recommendations • For regulated environments workflows using Human in loop. This work helps on understanding how GenAI can support decision making in risk environments while maintaining traceability. By combining my experience with hands-on AI engineering, I am studying on how past data help to produce consistent and high quality RCSA insights. Current Applied Projects Through various AI-driven mobile and productivity projects, I am exploring how NLP and LLM techniques can power natural-language reminders, contextual task automation, finance-aware reasoning, and early concept prototypes for AI-supported learning. These skills help me to combine enterprise-level engineering methods and hands-on experimenting in applied GenAI. Hands-On Implementations – Personal Productivity Platform I have implemented real world scenarios to my productivity applications that connects NLP, and training LLM with auto learning capability: • NLP-powered reminder extraction, handling natural language inputs, time normalization, recurrence logic, and smart task parsing • React Native UI and mobile workflows, including cross-platform design and authentication flows • NLP for financial tasks, such as parsing expense-related language and exploring early finance categorization models • Progressive AI implementation, adding features incrementally such as structured reminder extraction, tagging, and contextual understanding • Data-driven learning, experimenting with methods to train/update models using anonymized user data stored in the database These components form a foundation for an intelligent assistant capable of understanding tasks, reminders, schedules, and financial cues in everyday language. Engineering Work on Projects • Enabling NLP/LLM features in AI-ready mobile architectures • React Native, Expo, Node.js, and Firebase API-based • Comparing open source and custom models • UX design using AI Powered, making mobile experience focusing on conversional patterns. Technical Interests • PEFT, LoRA, QLoRA and reward modeling • Distributed training (Colab, cluster setups) • Quantization, pruning and knowledge distillation • On-device inference and mobile model optimization • Event-driven contextual AI agents • AI Focusing on people for learning, productivity and everyday tasks. Vision for the Long-Term Goal is to implement domain specific Language modals that combine understanding process, and contextual inference. These LLMs will power smart mobile and business apps that are clear, reliable, and easy to use. Interested in studying AI in both theoretically and practically. For example, I am more interested in learning transformer structures with great details and test those concepts using real data, behavior, and contexts.