Ambica Ghai

Led framework design by architecting persona-based multi-agent GenAI pipelines leveraging LangChain and RAG for enterprise voice and text analytics. Mentored AI engineers on LLM design, accelerating integration of intelligent agents into production-grade pipelines. Implemented structured retrieval and response models for multimodal datasets, improving generation accuracy by 30% in production systems.

Delivered 100+ hours of targeted AI, ML, and Power BI courses for 20+ industrial professionals per batch. Supervised research for 1 Ph.D. candidate and 3 graduate interns. Developed curriculum on AIML, AI Strategy, and Data Visualization integrating hands-on exercises with no-code tools.

Developed 4 hours of Coursera specialization material on applying ML in marketing. Implemented internship tracking system for 1000+ students, boosting operational efficiency. Delivered 100+ hours of lectures on Information Systems and Technology.

Project: Machine Learning Framework for Image Forgery Detection. Reduced fake content by 10% on social media platforms by designing and deploying robust image classification models leveraging CNNs. Achieved 90% accuracy in detecting manipulated images through rigorous training, testing, and transfer learning techniques using TensorFlow. Enhanced interpretability by creating data visualizations using Python libraries. Project: Sentiment Analysis Framework for Social Media Data. Developed a DNN integrating NLP and computer vision, delivering 89% accuracy on benchmark sentiment analysis tasks for multimodal data. Optimized data processing workflows using AWS and PySpark for scalable model training and deployment. Project: Image Recommendation Algorithm Development. Engineered scalable recommendation systems by combining YOLOv9-based object detection with CLIP multimodal embeddings for improved contextual image matching. Designed hybrid text-image fusion and dense retrieval architecture, achieving Recall@1 of 92% and Recall@5 of 98%.

Collaborated with e-commerce teams to deliver Agile-driven customized web apps, improving user engagement. Leveraged advanced statistical analysis to enhance customer satisfaction by 15%. Created Power BI and Excel dashboards for actionable management insights, increasing reporting efficiency. Authored comprehensive BRDs and technical specs, reducing development cycle time by 35%.