1. Investigate the root causes of the lost-in-the-middle phenomenon in long-context LLMs, especially within RAG frameworks. 2. Develop and evaluate a lightweight, fine-tuning-free method to mitigate positional bias in retrieved contexts. 3. Benchma

Client name or a organization name can be changed as per the Geo or Country, however the parent organization will be the actual Global Client Name. We are require all the subsidiary company from the upstream pointing to Parent organization using AI.

We intend to develop a framework which is capable of answering questions given a snippet of code.

To develop Web-Based Solution For Breast Cancer Detection Using Machine Learning And Deep Learning

Identifying the malicious text & frauds from a incoming text or whats app messages from various sources.

Design a planning agent that can synthesize a pipeline from a natural language specification primarily targeting data engineer persona.

Feedback forms could be descriptive and sometimes difficult to read through all the document and see the result, here we will use generative AI to summarize the result with Positive or negative feedback score to make it easy to understand.

Any project document, HLD or a Contract document are quite a big and difficult to go through all the pages. Also difficult to search if we are looking for any specific answer from multiple sections. Hence requirement is to generate a summarize which

The objective of this research work is to find possible way of detecting finger movements with help of powerful sensors, which would thereafter be used to detect the sign language that was made and then convert it to audio.