I’m Simerjit, an M.Sc. Artificial Intelligence student at Nanyang Technological University (NTU), Singapore, with a strong focus on Generative AI, LLMs, Multi-Agent Systems, Neurosymbolic AI, NLP, and Computer Vision.

Previously, I worked at NielsenIQ as a Research Analyst & Team Lead, delivering FMCG insights for Reckitt, Nestlé, and other Fortune 500 clients across multiple markets. That experience shaped how I think about data, experimentation, and AI systems that solve real business problems.

Turning data into decisions, AI into advantage and ideas into innovation.

Neurosymbolic AI for Business Intelligence
Hybrid framework using SenticNet + deep learning to analyse McDonald’s Singapore reviews and map them into actionable business themes.

FIBO Multi-Agent GenAI System
A 5-agent pipeline that turns text prompts into product designs and marketing assets with JSON-native workflows and validation loops.

Hardware-Aware U-Net for Stable Diffusion
Optimising Stable Diffusion v1.5 for edge deployment with activation sparsity, channel pruning, and improved FID.

CodeLah! (Code ↔ Natural Language Translation)
LoRA-fine-tuned CodeGen-350M with AST parsing and semantic validation for robust NL↔code translation.

AMEX Dynamic Rewards GenAI Platform
Built a GenAI-powered rewards personalization system with a Safe Action Engine (Policy-as-Code) and RAG-based engagement analytics for American Express.

FinFriend – TE Connectivity x AWS Hackathon
Designed an AI-powered financial guidance assistant using AWS Bedrock, LangGraph multi-agent orchestration, Prophet, RandomForest, and FAISS-based RAG, supporting 13 Indian languages.