PhD on contextualizing generative AI systems in structured data (m/f/x) Undefined

Interested in making representation and generative learning work with structured data (e.g. tables
in spreadsheets and databases) to accurately, securely, and efficiently, democratize insights from
data? This 4-year PhD starting September 2024 is for you!

Job description

Goal of the DataLibra project
Approximately 120 zettabytes of data has been collected worldwide but less than 1% is actually
used. Structured data, e.g. tables, spreadsheets, and relational databases, is prevailing in
organizations and typically informs important decisions in healthcare, governments and finance.
Yet, while AI has demonstrated a high impact on applications on text and images, proportional
progress on structured data is lacking. With the DataLibra project, we aim to close this gap, by
developing AI models and tools for structured data (Table Representation Learning), to help
organizations, of any size, domain, and level of data literacy, get insights from structured data,
efficiently, accurately and securely.

Goal of this PhD project
Following recent developments in AI, large language models (LLMs) have been explored for data
analytics tasks (e.g. text-to-sql), but show limited accuracy in domain-specific contexts with
structured data. In this project, we will investigate and design interactive intelligent systems for
data analytics tasks, while accounting for two key challenges: trustworthiness of the outputs
(factuality), and security constraints of proprietary data contexts as in healthcare, enterprises, and
governments. Potential directions to explore are agentic systems, retrieval augmented generation,
(instruction) fine-tuning, and others.

What you will be doing
 Inform a research agenda on the PhD topic for a timespan of four years.
 Develop methods and systems for contextualizing generative AI for analytics over
structured data.
 Actively collaborate with other researchers in the DataLibra project (students, 4-5 PhDs,
postdocs, PI) and external collaborators (e.g. Amsterdam UMC, University of Amsterdam).
 Communicate research outcomes through papers and presentations at conferences,
workshops and other (scientific) gatherings.
 Assist in relevant teaching activities at universities, such as thesis supervision and assisting
in courses.

Requirements

• A master (or equivalent) in computer science, artificial intelligence or similar.
• Self-drive, creativity, rigor, ownership, and excitement to push research in TRL forward.
• Theoretical knowledge of, or experience with, data management, human-computer

interaction, and/or machine learning such as representation and generative learning.

• Programming skills, e.g. Python, Java, or C++.
• Excellent command in English, verbal and written.
• Prior experience as a research assistant during (under)graduate studies is advantageous

Terms and conditions

The terms of employment are in accordance with the Dutch Collective Labour Agreement for Research Centres (“CAO-onderzoeksinstellingen”). The initial labour agreement will be for a period of 18 months. After a positive evaluation, the agreement will be extended by 30 months. The gross monthly salary, for a PhD student on a full time basis, is € 2,781 during the first year and increases to € 3,562 over the four year period. Employees are also entitled to a holiday allowance of 8% of the gross annual salary and a year-end bonus of 8.33%. CWI offers attractive working conditions, including flexible scheduling and help with housing for expat employees.

Please visit our websites for more information about our terms of employment:
https://www.cwi.nl/jobs/terms-of-employment and
https://www.nwo-i.nl/en/working-at-nwo-i/jobsatnwoi/

Information and application

About the PI
Madelon Hulsebos is an incoming tenure track researcher at CWI in Amsterdam and postdoctoral
fellow at UC Berkeley. She obtained her BSc and MSc from TU Delft, and her PhD from the
University of Amsterdam for which she did research at MIT and Sigma Computing. Her general
research interest is on the intersection of data management and machine learning, currently
focusing on Table Representation Learning. Madelon founded the Table Representation Learning
workshop at NeurIPS, and leads various related scientific efforts. She was awarded the BIDS-
Accenture fellowship for her postdoctoral research on retrieval systems for structured data at UC
Berkeley, and more recently the AiNed fellowship funding her 5-year DataLibra project at CWI.

What we offer
The terms of employment are in accordance with the Dutch Collective Labour Agreement for
Research Centres (“CAO-onderzoeksinstellingen”). The initial labor agreement will be for a period of
18 months. After a positive evaluation, the agreement will be extended by 30 months. The gross
monthly salary, for a PhD student on a full time basis, is € 2,781 during the first year and increases
to €3,562 over the four year period. Employees are also entitled to a holiday allowance of 8% of the
gross annual salary and a year-end bonus of 8.33%. CWI offers attractive working conditions,
including flexible scheduling and help with housing for expat employees.

Please visit our websites for more information about our terms of employment:
https://www.cwi.nl/jobs/terms-of-employment and
https://www.nwo-i.nl/en/working-at-nwo-i/jobsatnwoi/

Questions?
If you have questions about this position, please contact the Principal Investigator (PI): Dr. ir.
Madelon Hulsebos, madelon@berkeley.edu.
For more information about CWI, please visit www.cwi.nl or watch our video about working at CWI.

Interested? Apply!
If you are interested in pioneering research in TRL and the above profile fits you, we look forward to
receiving your application. Please use the “apply” button below to apply. We accept applications
until 21 July 2024.

Applications should include the following information in a single PDF file.
 Letter motivating your application for this position (self-authorship is appreciated).
 Curriculum Vitae, including potential publications, projects you worked on, etc., if any.
 A link to your (under)graduate thesis.
 Complete grade record of (under)graduate courses.
 The names and email addresses of two (academic) references.

A select number of applications will be invited to the first interviews to be planned between 25 and
31 July, selected candidates will get a scoped take-home assignment to be discussed in a second
interview. Please note that this position will start in fall 2024, if that does not fit your timeline, there
will be more PhD (and postdoc) openings between spring 2025 and spring 2026. You can leave your
email here if you would like to be informed of new openings

Diversity code

CWI encourages a diverse workforce: we endeavour to develop talent and creativity by bringing people from different backgrounds and cultures together. We recruit and select based on capabilities and talent. We strongly encourage everyone with the appropriate qualifications to apply for the vacancy, regardless of age, gender, origin, sexual orientation or physical abilities.

About CWI

Centrum Wiskunde & Informatica (CWI) is the Dutch national research institute for mathematics and computer science and is part of the Institutes Organisation of the Dutch Research Council (NWO). The mission of CWI is to conduct pioneering research in mathematics and computer science, generating new knowledge in these fields and conveying it to trade, industry, and society at large.

CWI is an internationally oriented institute, with 160 scientists from approximately 27 countries, an informal atmosphere and short lines of communication. We have an activity committee that organizes after-work activities and an informal women’s network.

CWI is located at Science Park Amsterdam, the home of AMS-IX, that is presently developing into a major location of research in the physical sciences in The Netherlands, housing the sciences of the University of Amsterdam as well as several other national research institutes next to CWI.

About Research group

.https://www.cwi.nl/en/groups/database-architectures/