Hey. Hi everyone. I am Sandid Gini, a technical program manager with over 17 years of experience right now working in Hitachi Digital Services here at Dallas. In my previous organizations, I have worked acro, I have worked in large technology transformation banking projects across multiple geographies. Today I'm here to talk about LLM Power Banking Organization coming to, coming back to the slides. The first slide, I'm excited to sh share with you a groundbreaking approach to one of the bank's banking's most persistent challenges, LLM Powered Banking Modernization. Coming back to the next slide, when we look at the banking industry today, we are facing that, what I call legacy burden. This pie chart tells us a stock story. Legacy maintenance is consuming up to 65% of IT budgets in most financial institutions. Think about that for a moment. Two thirds of your technology spend just keeping the lights on. Most banking systems still run on cobol, a language developer before many of us were born. It's reliable, but it's becoming increasingly difficult to maintain. The talent Pool of COBOL developers is shrinking. While demand for new digital services exponentially grown, I've seen this firsthand while working on transformation projects across multiple banks across multiple geographies. One CIO told me we are spending so much maintaining our legacy system that we can't invest in our innovation, our customer's demand coming to the slide three, it's a modernization. It's talking about the modernization paradox. So why not just modernize? The statistics on this slide, we, what they call the modernization paradox, traditional transformation approaches. Have an 85% failure rate projects that exceed budget missed deadlines, or simply fail to deliver and expected value. The average duration stretches beyond 24 months, and I have personally witness asked budget overturns of three to five times the original speeds. One project I consulted on began with an 18 month timeline and ended up taking over four years to complete. The reality is that financial institutions need a fundamentally new approach. Of all methods, ofri rip and replace man or manual code conversion simply cannot scale to meet modern demands, especially given the criticality of these systems and zero tolerance for errors. Coming back to the next slide, a new paradigm. This is where our new framework comes in. We have doubled three trial approach that SSEs the core challenges of banking modernization. At the foundation, we have a human AI partnership. This isn't about replacing human expertise, it's about amplifying it. Banking domain experts work alongside AI to ensure that business rules and regulatory requirements are preserved. Amelia Tire focuses on accelerated modernization. By using LLMs, we can transform code faster and more reliably than manual approaches, drastically reducing project timelines at the top. We enable competitive advantage through innovation capabilities. That were previously locked behind Legacy Constraints One Bank using approach was able to reduce their future deployment time from months to days. Our framework maintains regulatory compliance while transforming legacy systems a critical requirement in the highly regulated banking environment. Coming at the slide five, the results we're seen are truly transformative. This circular diagram highlights the key metrics. From our earlier implementations, we have achieved an 80% reduction in human effort for code trans translation Tasks that would take developers weeks can now be completed in days testing. Notification cycles are 60% faster. As our approach allows for automated equivalence testing between all the new systems post implementation, we are seeing 75% fewer regression defects. This is crucial for maintaining the stability that banking customers expect. Perhaps most compelling to execute is a 35%. Lower total cost of ownership. One regional bank I worked with projected the savings of over $12 million or five years through this approach. These aren't the theoretical projections. They are the results that early adopters are already experiencing. Coming back to the technical architecture, this slide outlines the four stage process that powers our approach. First, legacy code ingestion. Handles COBOL passing and domain mapping. Next LLM translation applies banking specialized model processing to the code. Third verification ensures background and behavior equivalence through rigorous testing. Finally, human review provides expert validation refinement. This framework successfully overcomes key challenges while handling banking specific complexities. LLM coming back to the slides. Next slide is LLM Innovation. The technical innovations driving this solution are threefold. Extended context windows allow models to process entire banking module simultaneously. Previous reserving relationship between functions and improving translation accuracy. Financial domain, fine tuning, ensures model understandings, accounting principles, transaction processing and regulatory requirements, multi-language processing. Our proficiency enables. A seamless translation between cobol, Java, Python, and other modern languages while maintaining semantic equivalence. Coming back to the next slide. It's a human oversight system. Despite power of ai, human expertise, domains essential. Our system incorporates role-based reviews, where engineers, compliance offices, and domain exploit validate different aspects of the transformation. Automated compliance flagging identifies potential regulatory issues for human review. Performance verification ensures equivalent behavior through side by side runtime. Finally, feedback loop integration allows human corrections to continuously improve the AI model. Coming back to the next slide, investment impact. This power shot demonstrates the compelling business case. For this approach, we see dramatic predictions and maintenance costs feature time to market and technical debt index. After implementation, these real role metrics confirm that reduced cost and acceleration innovation. Creating lasting complete competitive advantages for adopters. Coming back to the next slide this is the transaction system where we wanted to show the case study. A major bank transformed this code transaction system processing 3 million daily transactions from 10 million lines of COBAL code, which is a 30-year-old. The process followed our framework, legacy assessment, preparation, phase IT translation, and finally successful migration to containerized microservices with improved performance and its scalability. This case proves the varied viability of our approach, even of for the most complex infrastructure. Coming back to the implementation roadmap for our organization looking to adopt this approach, we need to provide a clear four phase roadmap. Beginning with the assessment of preparation. The process moves to pilot transformation of non-critical modules, 10 to scale implementation across larger systems and financially. And finally, continuous improvement through ongoing refinement, this actionable path. Forward minimizes risk while maximizing the benefits of AI powered modernization. Coming back to this next slide, transform the banking future. This is the closing slide that emphasizes the three key benefits, reducing technical gap by transforming legacy systems from liability to advantage. Accelerate innovation by delivering new features in days instead of months. Enhance compliance through improved regulatory. And automated checks. This the time for AI powered modern Magnetization is now offering a way to transform legacy systems while preserving business logic and minimizing the risk. Any questions? Thank you so much.