Extends Indic reasoning benchmarks with bias injection and Chain-of-Thought cause-effect generation. Enables robust evaluation of LLM trustworthiness and reasoning capabilities (GRPO/DPO styles).

A semi-anonymous doubt solving platform featuring 4 AI study agents. Built with scalable async microservices, generating personalized content using pluggable LLM adapters.

Novel architecture combining BART Encoders with ViT backbones to detect sarcasm in image-text pairs. Based on 'Target-Augmented Shared Fusion-based Multimodal Sarcasm Explanation Generation'.

Scalable creator economy platform with microservices. Features real-time chat, NFT marketplace, and secure transactions. Demonstrates complex backend orchestration.