Shikhar Verma

Designed and implemented a real-time alert framework using Server-Sent Events (SSE) and Kafka, reducing supervisor approval time by over 50% for critical payment transactions.

Developed and integrated Generative AI NLP and LLM-powered features for branch teller transactions, reducing filtering time by 35% and increasing user efficiency by 35% through error recovery assistance.

Engineered a chatbot-integrated solution for end-to-end Branch Teller transactions, decreasing transaction time by 40%, and designed a Cash Prediction use-case to enhance operational efficiency.

Automated Teller transaction testing, boosting testing efficiency by 35%, and continuously improved code coverage by 25%, while actively resolving support production bugs to maintain stability.

Implemented a Charge Benefits feature for customers, successfully reducing total charges based on their relationship with the bank.

Built and deployed an Oracle Banking Origination Chatbot using Oracle Digital Assistant (ODA), significantly reducing the time required to initiate applications for savings/current accounts, loan, and credit card accounts by 60%.

Designed the comprehensive conversational flow of the speech-enabled ODA chatbot utilizing YAML language to ensure intuitive user interaction.

Developed and integrated custom NodeJS components for robust backend functionality, enabling seamless data exchange and process automation for the chatbot.