Eligibility: Master's degree in Computer Science, AI, NLP, Machine Learning, Data Science, Computational Linguistics or a closely related field.
Language: English at C1 level; no German requirement is stated.
Deadline:14 November 2026; start as soon as possible.
Table of Contents
The German National Library of Science and Technology (TIB) in Hannover is offering a paid doctoral research position focused on Natural Language Processing, causal AI and scientific reasoning.
The position is part of the Leibniz-funded EcoAIRE project, which aims to develop AI systems capable of assessing scientific hypotheses from fragmented and heterogeneous research evidence.
You will work on methods that transform scientific publications, datasets, images and other research artefacts into structured representations of:
- hypotheses
- variables
- context
- uncertainty
- evidence
- causal relationships.
The role combines large language models, agentic AI, causal reasoning, Semantic Web technologies and scientific knowledge representation.
The position is paid at TV-L E13, normally at 100% employment with 39.8 hours per week, and is initially limited to three years. Part-time employment is generally possible.
Applications are open until 14 November 2026.
TIB NLP and Causal AI PhD at a Glance
Information
Details
Position
Research Associate / Doctoral Researcher — NLP and Causal AI for Scientific Reasoning
Ready to Start Your German Study Adventure?
Explore our university finder tool to discover the perfect German university for your academic goals and preferences.
Naman Kumar has lived in Germany for over eight years and has spent more than five years helping international students navigate German universities, courses, scholarships, and admissions. Having studied in Germany himself and received scholarship support, he combines first-hand experience with practical guidance. He has supported more than 200 students in finding suitable study and funding opportunities in Germany.
Subscribe to Our Newsletter
Get the latest updates on studying in Germany, visa changes, and scholarship opportunities.
By subscribing, you agree to our Privacy Policy and consent to receive updates from our company.
Institution
German National Library of Science and Technology (TIB)
Association
Leibniz Association
Location
Hannover, Germany
Project
EcoAIRE
Research group
Open Research Knowledge Graph, Data Science and Digital Libraries
Academic level
Doctoral
Salary grade
TV-L E13
Employment
100%
Weekly hours
39.8
Part-time
Generally possible
Contract
3 years
Funding condition
Subject to final Leibniz funding approval
Start
As soon as possible
English
Minimum C1
German
No requirement stated
Reference number
44/2026
Deadline
14 November 2026
TIB is one of Germany's major information and research institutions and combines scientific infrastructure with active research in digital libraries, knowledge graphs, artificial intelligence and open science.
What is this TIB PhD about?
The doctoral position focuses on developing AI systems that can do more than simply extract individual facts from scientific literature.
The goal is to build systems that can understand how scientific evidence fits together and evaluate:
- which hypotheses are supported
- nwhich variables matter
- under which conditions a claim holds
- how uncertainty affects conclusions
- how causal relationships can be represented and tested.
The research therefore sits at the intersection of: NLP + LLMs + causal AI + scientific knowledge representation.
Rather than treating research papers as isolated text documents, the project aims to build structured and machine-actionable representations of scientific reasoning.
What is the EcoAIRE project?
The position is part of: EcoAIRE — AI-assisted hypothesis assessment in invasion biology and the restoration of freshwater ecosystems under fragmented and heterogeneous ecological evidence.
EcoAIRE brings together expertise from three major German research organisations.
TIB
TIB contributes expertise in:
- natural language processing
- scientific knowledge representation
- Open Research Knowledge Graph technologies.
Leibniz Institute of Freshwater Ecology and Inland Fisheries
IGB contributes expertise in:
- freshwater ecology
- invasion biology
- ecosystem restoration.
German Research Center for Artificial Intelligence
DFKI contributes expertise in:
- causal modelling
- causal inference
- artificial intelligence.
The project therefore combines computer science with real scientific evidence from ecology.
What will you research during the PhD?
Your work will focus on several interconnected research problems.
Scientific information extraction
You will develop methods that extract and represent:
- scientific hypotheses
- variables
- contextual conditions
- evidence
- uncertainty
- causal relationships.
The goal is not just conventional named-entity or relation extraction.
You will investigate how individual pieces of evidence can be connected into a larger model of scientific reasoning.
Multimodal scientific knowledge integration
Scientific evidence does not exist only in article text.
You will work on a multimodal and provenance-aware framework that can integrate sources such as:
- research publications
- images
- supplementary information
- datasets
- other research artefacts.
The resulting representations should support automated hypothesis assessment.
Large language models and agentic AI
You will develop language-model-based and agentic methods for:
- extracting information
- normalising scientific concepts
- aligning information across sources
- synthesising evidence
- representing uncertainty
- identifying causal or mechanistic relationships.
The vacancy explicitly expects practical experience with:
- transformer models
- large language models
- agentic or multi-agent workflows.
Causal AI and scientific reasoning
One of the most distinctive parts of this PhD is the connection between NLP and causal reasoning.
You will investigate approaches inspired by:
- causal models
- world models
- simulations
- intervention-based reasoning.
The aim is to assess how scientific hypotheses behave under changing conditions rather than simply identifying whether a paper contains a claim.
What does “scientific reasoning” mean in this PhD?
Scientific reasoning here means modelling how scientific evidence supports or challenges hypotheses.
For example, an AI system may need to understand
:- what hypothesis a study tests
- which variables influence the outcome
- what experimental conditions apply
- how strong the evidence is
- what uncertainty remains
- whether the result generalises to another context.
The EcoAIRE project aims to build AI methods capable of reasoning across these relationships.
Will you work with LLMs?
Yes. Hands-on experience with transformer-based models and large language models is explicitly required.
The vacancy mentions frameworks such as:
- PyTorch
- Hugging Face Transformers
or comparable technologies.
You should be able to demonstrate practical experience using modern language models for NLP applications.
Do I need experience with agentic AI?
Yes. The position specifically requires practical experience developing agentic or multi-agent AI workflows.
Relevant experience can include:
- tool use
- structured outputs
- agent orchestration
- iterative reasoning pipelines
- multi-agent collaboration patterns.
This is therefore not a traditional NLP PhD focused only on classification or information extraction.
Do I need RAG experience?
RAG experience is desirable, but not listed as mandatory.
The vacancy specifically values experience building advanced retrieval-augmented generation pipelines using technologies such as:
- Sentence Transformers
- LangChain
- LlamaIndex
- FAISS
- Qdrant
- embedding-based retrieval
- reranking
- vector search
- structured generation.
If you have built production or research RAG systems, highlight that clearly in your application.
Do I need causal inference experience?
Not necessarily. Knowledge of causal modelling and causal inference is listed as a desirable qualification rather than a core mandatory requirement.
Relevant knowledge could include:
- causal graphs
- structural causal models
- intervention-based reasoning.
Strong NLP, LLM and agentic-AI experience may therefore still make you competitive even if causal inference is a newer area for you.
Do I need Semantic Web experience?
No, but it is desirable. TIB explicitly lists experience with Semantic Web technologies as an advantage.
This is understandable because the position sits within TIB's Open Research Knowledge Graph and Data Science research environment.
Experience with technologies such as:
- RDF
- knowledge graphs
- ontologies
- SPARQL
- linked data
could therefore strengthen your profile, although the vacancy does not require a specific Semantic Web technology.
Do I need an ecology background?
No. Domain expertise in ecology is listed as desirable, not mandatory.
Relevant ecological areas include:
- freshwater ecology
- invasion biology
- restoration ecology.
You can still be a strong candidate with a computer-science or NLP background and no formal ecology degree.
The project is explicitly interdisciplinary, so domain expertise can also be developed through collaboration with the ecological research partners.
Who can apply?
You need a completed Master's degree or equivalent in Computer Science or a closely related field.
The vacancy explicitly gives examples including:
- Artificial Intelligence
- Natural Language Processing
- Machine Learning
- Data Science
- Computational Linguistics.
Closely related technical fields may also qualify if your academic and research background fits the position.
Can I apply with a Data Science Master's?
Yes. Data Science is explicitly named among the relevant academic backgrounds.
You should still demonstrate the required practical experience in:
- Python
- LLMs or transformers
- software engineering
- agentic AI workflows.
Can I apply with a Computational Linguistics degree?
Yes. Computational Linguistics is explicitly listed as a suitable related field.
Applicants from computational linguistics should emphasise technical NLP experience, particularly with modern transformer models and Python-based research software.
Can I apply with an AI or Machine Learning Master's?
Yes. Artificial Intelligence and Machine Learning are explicitly listed as relevant academic backgrounds.
Your application will be particularly relevant if your previous research includes:
- NLP
- LLMs
- knowledge graphs
- causal reasoning
- scientific information extraction
- RAG
- agentic AI.
Is previous research experience required?
The vacancy requires demonstrable experience in research or software development using Python.
It does not publish a required number of years.
That means you do not necessarily need several years of full-time industry experience.
Relevant evidence could come from:
- Master's research
- research assistant work
- academic projects
- software projects
- internships
- open-source contributions.
What Python experience do I need?
The position requires practical research or software-development experience using Python.
TIB also expects familiarity with modern software-engineering practices such as:
- Git-based collaboration
- automated testing
- reproducible experimentation.
This means the role expects more than notebook-only experimentation.
You should be comfortable building reproducible research software.
Which AI technologies are relevant?
The official vacancy names or references technologies including:
- Python
- PyTorch
- Hugging Face Transformers
- Sentence Transformers
- LangChain
- LlamaIndex
- FAISS
- Qdrant.
The position also involves:
- transformer-based models
- LLM fine-tuning or adaptation
- RAG
- embeddings
- reranking
- vector search
- structured generation
- agentic workflows
- causal models.
You do not need to know every listed library, but your experience should demonstrate strong modern NLP and AI capabilities.
Do I need German?
No German requirement is stated in the official vacancy.
The required language is English at minimum C1 level.
English is required for:
- scientific communication
- collaboration
- publication.
The job advertisement does not list a German CEFR requirement or German-language certificate.
Do I need IELTS or TOEFL?
The official vacancy specifies English at least C1 but does not state that you must submit a specific IELTS, TOEFL or Cambridge certificate.
Do not assume a particular test score is mandatory unless TIB requests proof during the application or hiring process.
Can international applicants apply?
Yes. TIB explicitly states that it welcomes applications regardless of:
- gender
- nationality
- ethnic and social origin
- religion or belief
- disability
- age
- sexual orientation and identity.
International applicants therefore can apply, provided they meet the academic, technical and language requirements.
What happens if my Master's degree is from outside Germany?
Applicants with foreign university qualifications need to pay particular attention to German degree-recognition requirements.
The vacancy states that a Statement of Comparability from the Central Office for Foreign Education (ZAB) is mandatory for the final assessment of hiring requirements during the hiring process for foreign university degrees.
A fee is charged for the ZAB assessment.
This requirement relates to the formal hiring process rather than the scientific evaluation of your application.
Do I need the ZAB certificate before applying?
The vacancy says that ZAB comparability evidence is mandatory during the hiring process for the final assessment of eligibility where a foreign university degree is involved.
It does not explicitly state that every foreign applicant must already possess the completed ZAB document on the day they submit the initial application.
If your foreign degree is clearly recognised in Germany through ANABIN, check the current TIB hiring instructions and German recognition rules for your case.
Is this PhD paid?
Yes. The position is salaried according to TV-L E13.
This is an employee doctoral position rather than a scholarship-only doctorate.
The official Leibniz vacancy does not state one fixed individual monthly salary, because TV-L pay depends on the applicable experience step and employment conditions.
Is this a full-time PhD?
The standard position is 100% with regular weekly working time of: 39.8 hours.
The position is generally also suitable for part-time employment.
Can I work part-time?
Yes. TIB explicitly states that the position is generally suitable for part-time work. The exact part-time arrangement would need to be agreed with the employer.
Is remote work possible?
Yes, to a limited extent. The vacancy states that mobile work is possible to a certain extent subject to operational requirements.
This is therefore not advertised as a fully remote doctoral position.
The primary workplace is Hannover.
How long is the contract?
The contract is initially three years.
However, there is one important condition the appointment is subject to final funding approval by the Leibniz Association.
Applicants should understand this as a formal funding condition attached to the position.
When does the PhD start?
The advertised start date is as soon as possible.
No exact calendar start date is published.
If selected, the final start date would therefore need to be agreed during the recruitment process.
Will I publish research?
Yes. Publishing research is explicitly part of the role.
You will be expected to publish findings in peer-reviewed:
- NLP conferences and journals
- Semantic Web venues
- AI venues
- machine-learning conferences and journals.
You will also contribute to:
- project reports
- dissemination activities.
How will the AI systems be evaluated?
The project includes research on multifaceted evaluation frameworks for AI-supported scientific hypothesis assessment.
The vacancy specifically mentions:
- expert-grounded evaluation
- LLM-as-a-judge approaches.
This means evaluation methodology itself is a research topic rather than merely a final benchmarking step.
What research infrastructure will you use?
TIB states that the researcher can build on several existing research infrastructures and software components, including:
- TIB AIssistant
- Open Research Knowledge Graph (ORKG)
- ORKG Ask
- software from the SciKnowOrg ecosystem.
This provides an existing technical ecosystem rather than requiring the doctoral researcher to build everything from scratch.
What research group will you join?
The position belongs to Program Area D — Open Research Knowledge Graph, Data Science and Digital Libraries Research Group under Prof. Dr. Sören Auer.
The research environment focuses on areas such as:
- Open Research Knowledge Graphs
- scientific knowledge organisation
- digital libraries
- AI and data science.
The vacancy also identifies Dr. Jennifer D'Souza as the contact person for questions about the position.
What benefits does TIB offer?
TIB lists a substantial public-sector employment package.
Benefits include Annual special payment
Employees receive a special annual payment under the applicable public-service conditions.
30 days of annual leave
You receive 30 days of vacation for a five-day working week.
TIB also provides additional days off on:
- 24 December
- 31 December.
Public-sector pension
The position includes VBL supplementary pension for public-sector employees.
Training and professional development
TIB offers internal and external:
- professional training
- further education
- development opportunities.
Research travel and conferences
The position includes funding for:
- necessary research equipment
- conferences
- research visits.
Flexible work
TIB offers:
- flexitime
- mobile work
- remote-working options where operationally possible
- family-friendly working arrangements.
Health and university benefits
Employees can access:
- workplace health-promotion programmes
- discounted University Sports Hannover
- discounted meals in Studentenwerk Hannover canteens.
What documents do you need?
The vacancy asks applicants to submit complete application documents.
However, the public job posting does not provide a definitive item-by-item checklist such as:
- mandatory motivation letter
- mandatory transcript
- mandatory references
- mandatory publication list.
You should therefore not present any such item as compulsory unless it appears in the online application form.
For email applications, TIB requests the application as one PDF file with a maximum size of: 10 MB.
Should I include salary expectations?
TIB states that applicants are invited to provide their annual salary expectations.
However, the position itself is already classified at TV-L E13.
This means your actual remuneration is still governed by the public-sector salary framework and applicable experience step.
How to apply for the TIB NLP and Causal AI PhD
Step 1 — Check your academic background
You need a Master's or equivalent in Computer Science or a closely related area.
Step 2 — Evaluate your technical fit
Make sure you can demonstrate:
- Python experience
- modern software engineering
- transformer or LLM experience
- agentic or multi-agent workflows
- C1 English.
Step 3 — Highlight preferred expertise
If applicable, highlight experience with:
- RAG
- LLM fine-tuning
- causal inference
- Semantic Web
- ecological research.
Step 4 — Prepare your complete application
Follow the online portal requirements and include all requested documentation.
Step 5 — Address foreign-degree recognition if necessary
If your degree was obtained outside Germany, be prepared to provide the required ZAB comparability evidence during the hiring process.
Step 6 — Apply online
TIB provides an official online application route for vacancy: 44/2026.
Step 7 — Alternative email application
You can also submit a single PDF of no more than 10 MB to:bewerbung@tib.eu
Step 8 — Submit before the deadline
Your application must arrive by: 14 November 2026.
Application Deadline
The application deadline is: 14 November 2026
The vacancy was published on: 16 October 2026.
The position is expected to begin:as soon as possible.
There is no separate reference-letter deadline stated in the public vacancy.
Planning to study at a German University?
Explore German universities and degree programmes and compare your study options before applying.
