Nilay Sharma

Architected and delivered a MuleSoft API integration that automated Salesforce and AWS Marketplace private offers, directly enabling $50 million in quarterly sales and streamlining deal management for Account Executives.

Led the end-to-end development lifecycle from API design (RAML) to CloudHub 2.0 deployment, ensuring system reliability through MUnit test cases and active monitoring.

Engineered a robust, API-led MuleSoft solution connecting Salesforce with AWS Marketplace, implementing asynchronous webhook processing and real-time notifications to ensure data integrity.

Developed an AI-powered chatbot using Gemini APIs integrated with Salesforce for context-aware user interactions, and built a knowledge graph-powered recommendation system using RAG, FastAPI, Docker, and GCP Cloud Run.

Transformed legacy Visualforce/Sencha UIs into LWC, reducing page load time by 40% and significantly improving user experience across key accounting modules.

Built a scalable audit file generator supporting 15-25 million records (an 8x increase from 2-3 million) for various countries, leveraging a dynamic SAQL engine and metadata-driven filters.

Implemented a Bryntum grid in LWC, enabling high-performance rendering and editing of 2,000+ line items per invoice record for mass updates and error resolution.

Led cross-functional efforts in refining epics and stories, collaborating with management, UI/UX, developers, and tech writers to deliver seamless solutions.

Developed generic Lightning components and loosely coupled backend using a factory design pattern, reducing code duplication and maintenance efforts.

Developed and trained a novel deep learning model (Transformers, LSTMs, self-attention) for automated video interview analysis, boosting correlation between AI and human evaluations by 62% (from 0.4 to 0.65).

Executed the end-to-end machine learning workflow, encompassing data cleaning, feature extraction from raw video/audio, and model validation for assessing soft skills like confidence and engagement.