Charvi Jain

I am a second year doctoral student in the Faculty of Computer Science, TU Dresden under the supervision of Prof. Jens Lehmann and Prof. Ivo Sbalzarini. My research focuses on knowledge-grounded large language models, investigating how conversational AI systems can dynamically retrieve and effectively integrate external scientific knowledge with parametric knowledge to produce long-form reports. I am broadly interested in autonomous, agentic architectures for literature synthesis and hypothesis generation, with the overarching goal of building trustworthy AI systems that meaningfully accelerate scientific discovery.

Previously, I worked as an LLM Research Engineer at Fraunhofer IAIS for training 7B decoder based models from scratch with A100 GPUs under the OPENGPTx project.

I received my Master’s from TU Dresden, Germany with major in Deep Learning and my Bachelor’s from IIIT HP, India in Computer Science and Engineering. During the tenure of these degrees, I worked on a wide range of projects including machine learning, computer vision and tabular synthetic data generation.

I aim to work on projects that either push the boundaries of biology and medicine to improve lives or unlock new discoveries about the universe, ultimately making a lasting impact on humanity.