Postdoctoral Researcher – Large Language Models for Research Data Management in Physics (Physics-LLM)

Friedrich-Alexander-Universität; Bayern; Hybrid; EUR 54,000 - 70,000; Full time

Benefits offered by this job

30 days leave

Occupational pension

Onboarding support

Team activities

Public transport nearby

Childcare options

Training courses

Health management

Job summary

FAU in Erlangen-Nürnberg invites applications for a PhD-level role focusing on developing an LLM-enhanced Research Data Management toolkit within the Physics-LLM project. The work centers on metadata extraction, information retrieval, and system interfaces for APP and PaN communities.

You will collaborate with DESY, FZJ, and TUM to advance reproducible workflows and open-source infrastructure for research data management in fundamental sciences.

Qualifications

• Ph.D. in CS/Physics/Data Science or closely related field.
• Strong background in NLP, LLMs, and related techniques (RAG, fine-tuning, knowledge graphs).
• Familiarity with APP/PaN data contexts is desirable.
• Experience with software development and deploying computational models in research environments (Python).
• Proven ability to design rigorous experimental methodologies and benchmarking.
• Excellent communication to translate complex concepts for diverse audiences.

Responsibilities

• Establish a foundation for retrieving information from publications and data interfaces in Astroparticle Physics.
• Establish vocabularies and knowledge graphs, using LLMs to extract entities and relationships.
• Develop mechanisms to convert natural language queries into structured queries for data retrieval.
• Benchmark LLM performance in information retrieval using community datasets.

Skills

NLP

LLMs

RAG

Fine-tuning

Knowledge graph construction

Python

Education

Ph.D. in Computer Science, Physics, Data Science, or related field

Tools

Python programming

Software deployment

System architecture

Job description

Benefits: We Have a Lot To Offer

• Regular promotion to the next level and increase in salary pursuant to the collective bargaining agreement for the public service of the German Länder (TV-L) or remuneration pursuant to the Bavarian Public Servants Remuneration Act (BayBesG) plus an additional annual bonus
  
• 30 days annual leave at five working days per week with additional free days on December 24 and 31
  
• Occupational pension scheme and asset accumulation savings scheme
• Excellent support during the academic qualification phase
  
• Thorough onboarding process with a dedicated team
  
• Joint team activities
  
• Place of work within comfortable walking distance of public transport
  
• Family-friendly environment with childcare options, also during school holidays
  
• A wide range of training courses and opportunities for professional development
  
• Active health management

Your Tasks

Project Overview

The Physics-LLM project is a consortium dedicated to develop an LLM-enhanced Research Data Management (RDM) toolkit based on the FAIR principles. The primary goal is to overcome challenges in data collection, processing, storage, sharing, and finding within the physics domain by leveraging advancements in Natural Language Processing (NLP) and Large Language Models (LLMs).

The project is highly interdisciplinary, uniting researchers from various ErUM communities. The role will focus on developing and deploying LLM-based solutions for metadata extraction, information retrieval, and system interfaces, primarily serving the Astroparticle Physics (APP) and Photon and Neutron (PaN) communities.

Key Responsibilities

• Establish a foundation for retrieving relevant information from publications and data interfaces in Astroparticle Physics.
• Establish controlled vocabularies and knowledge graphs, leveraging LLMs to recognize entities, extract relationships, and semantically enrich data.
• Develop mechanisms to convert natural language user queries into structured queries for efficient data retrieval.
• Benchmark LLM performance in information retrieval tasks using community-specific datasets.

General Contributions

• Collaborate closely with team members from DESY, FZJ, and TUM to ensure seamless integration of research outputs across the consortium.
• Participate in the development of agentic systems (WP C.1) to enhance workflow design and data analysis reproducibility.
• Contribute to the documentation, benchmarking, and publication of project results in high-impact scientific venues.

Your Profile

Required Qualifications

• Education:Ph.D. in Computer Science, Physics, Data Science, or a closely related quantitative field.
• Technical Expertise:Strong background in Natural Language Processing (NLP), Large Language Models (LLMs), and related techniques (e.g., Retrieval-Augmented Generation (RAG), fine-tuning, knowledge graph construction).
• Domain Knowledge:Familiarity with the data structures, experimental workflows, and publication formats within Astroparticle Physics (APP) and/or Photon and Neutron (PaN) communities is highly desirable.
• Software/Systems:Experience with software development, system architecture, and deploying computational models in research environments. Proficiency in relevant programming languages (e.g., Python).
• Research Skills:Proven ability to design rigorous experimental methodologies, conduct quantitative benchmarking, and contribute effectively to complex, multi-partner scientific collaborations.
• Soft Skills:Excellent communication skills for translating complex technical concepts into actionable insights for diverse scientific audiences.

Additional Information

This position offers an unparalleled opportunity to work at the intersection of cutting-edge AI and fundamental physics research at a world-leading institution. You will be part of a consortium focused on creating sustainable, open-source infrastructure that will have a global impact on how research data is managed in fundamental sciences.

Notice

For all job postings, the following applies: Friedrich-Alexander-Universität promotes equal professional opportunities for women. Women are therefore expressly encouraged to apply.

Severely disabled persons as defined by the German Severely Disabled Persons Act will be given preferential consideration if they have the same professional qualifications and personal aptitude and if the advertised position is suitable for severely disabled persons. Details can be found in the respective job posting under “Remarks.”

At the applicant’s request, the Equal Opportunities Officer may be included in the job interview without any disadvantage to the applicant.

All advertised positions are generally suitable for part-time work unless stated otherwise in the job posting.