Prajvi Saxena

Researcher @Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI)

Saarbrücken, DE
MOBILE NUMBERS
+91 *********19

Signup · Get unlimited contacts

WORK HISTORY

Apr 2024 — Present

Researcher @Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI)

View department →

Saarbrücken, DE

Managing end-to-end processes from hypothesis formulation to machine learning cloud production, encompassing tasks such as data collection, cleaning, model development, and evaluation.•Computational efficient Large Language Models (LLMs): Developed and implemented ANON, an adaptive knowledge distillation framework using Quantized Low‐Rank Adaptation (QLoRA) to compress LLMs, reducing computational costs by 95%, while retaining high performance for domain‐specific applications (experimented with models such as LLaMa‐2, GPT‐2, and OPT).• Instruction Tuning LLMs: Developed as team leader of 8 researchers the Artificial Mental Model (AMM) for personalized patient care in knee rehabilitation. Utilized LLMs such as LLaMa‐3.1 (70B), and Mixtral 8×7B to create models that predict patient pain levels and optimize rehabilitation plans.

EDUCATION

N/A

Universität des Saarlandes

Master of Science - MS, Data Science and Artificial Intelligence.

N/A

Christ University, Bangalore

Bachelor of Technology, Computer Science

ABOUT PRAJVI SAXENA

I\'m an AI Researcher specializing in Large Language Models (LLMs) and Natural Language Processing (NLP). Currently, I’m part of the Smart Service Engineering (SSE) group at the German Research Center for Artificial Intelligence (DFKI), where we design smart service systems that address both technical and economic challenges, delivering industry-specific AI solutions.With over five years of professional experience at DFKI and a Master\'s degree in Data Science and Artificial Intelligence from Saarland University, I have a solid foundation in deep learning and machine learning. My expertise includes AI agentic framework, LLMs reasoning, computational efficient LLMs, Retrieval-Augmented Generation (RAG), multimodal models and domain-specific LLMs.I\'m passionate about using AI to solve real-world problems and dedicated to advancing the field through innovative research and collaboration. Always eager to connect with like-minded professionals and explore new opportunities to make a positive impact through AI.

This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.