Research Institute: Institute of Statistics, Biostatistics and Actuarial Sciences (ISBA), part of
LIDAM at UCLouvain.

Supervisor: Prof. Olivier Caelen.
Contract: Full-time employment contract with the industrial partner, fully funded for 48
months and organized as two successive 24-month periods.

Start date: October 2026.
Research setting: Approximately 50% at UCLouvain in Louvain-la-Neuve and 50% with the
industrial partner.

Scientific fields: Temporal causal inference, time series, machine learning, building energy
systems.

Proposed PhD topic

This applied PhD will develop temporal causal-inference methods for multivariate time series
from building sensor networks. Temporal causality is the scientific core: the project will
develop models of causal and dynamical relationships from complex, noisy data, incorporating
partial structural knowledge and domain expertise. The methods will be evaluated on their
ability to distinguish sensor faults from legitimate operational changes, detect physical
equipment malfunctions, and validate or reconstruct energy-flow graphs.
Methodological research, implementation, and validation on real-world industrial data will be
carried out with a partner active in building energy and property management.

Requirements:

• A Master's degree in statistics, computer science, engineering, data science, or a related
  field.
• Solid foundations in machine learning or statistical learning, time-series analysis, and
  good Python programming skills.
• Strong written and spoken English, and French proficiency at level B2 or higher
  (mandatory).
• Scientific rigor, autonomy, and interest in applied research. Experience in causal
  inference, dynamical systems, graphical models, or building energy systems is an asset.

We offer:

The selected candidate will join ISBA/LIDAM at UCLouvain, a collaborative research
environment combining statistics, machine learning, data analysis, and applied modelling.

• Full funding for a full-time PhD over 48 months, organised as two successive 24-month
  periods;
• A project combining methodological depth, real-world industrial data, and direct
  practical relevance;
• Regular interaction with academic and industrial experts;
• Opportunities for doctoral training, international conferences, and research collaborations.

Application procedure:

Interested applicants should send a single PDF file to: olivier.caelen@uclouvain.be
with the email subject: PHD CALL – [YOUR LAST NAME]
The application should include:

• A detailed CV;
• Transcripts for Bachelor's and Master's studies;
• A motivation letter (maximum one page);
• The Master's thesis, a research report, or another writing sample, if available;
• Links to code repositories, if available;
• Names and contact details of one or two academic references, if possible.

Application deadline: Thursday, 10 September 2026.

Interviews: Shortlisted candidates will be invited for an interview shortly after the application
review.

Originally from Côte d’Ivoire, Keivan Diakité has built a career at the intersection of applied mathematics, actuarial science, and research. After completing a PhD at LIDAM (ISBA), he is now a Data Scientist at Sigedis, where he directly contributes to the analysis and understanding of the pension system in Belgium.
His path reflects an increasingly common trajectory: that of a researcher applying his expertise to concrete societal challenges.
 

What motivated you to pursue a PhD, and why at LIDAM?

I chose LIDAM for its reputation, it is one of the most recognized actuarial research centers in Europe. But beyond prestige, it was mainly the intellectual environment that attracted me.
My PhD topic was interdisciplinary, which allowed me to explore several related fields and interact with experts from diverse backgrounds. This diversity was essential: it helped me step back, combine perspectives, and enrich my thinking.
 

“A PhD did not just deepen my knowledge—it broadened the way I think.”

How would you describe your PhD years?

They were shaped by a rather unique experience: I started my PhD shortly before the COVID-19 lockdown.
This was a significant challenge. For a large part of my early years, I could not attend conferences or academic events, which are typically essential in a doctoral journey.
But this period also taught me something fundamental:
the ability to adapt to uncertainty.

“Despite the distance, I managed to maintain excellent communication with my supervisor and colleagues. This was key to continuing my progress.”

What were the main challenges during that time?

Beyond isolation, the most difficult part was building a research dynamic without “natural” interactions.
In a PhD, many ideas emerge from informal discussions, seminars, and meetings. I had to recreate that environment remotely, in a more proactive way.
 

“This experience taught me to be autonomous, but also not to hesitate to actively seek feedback.”

How did the transition to your current role at Sigedis happen?

The transition was very smooth. My PhD research focused in particular on fairness in pension systems, a topic directly related to my current work in supplementary pensions.
In practice, I was able to reuse my research in an applied context, for instance through reports that contribute to discussions on pension reform.
The main difference lies in the pace:
the professional world is much faster and more results-oriented than academic research.
 

What does your current role look like in practice?

Within the Data Science team at Sigedis, I mainly work on the PensionStat.be platform, which is the reference for pension statistics in Belgium.
My role includes:
•    producing key indicators on statutory and supplementary pensions 
•    analyzing data to address institutional questions 
•    writing thematic reports for stakeholders such as the Court of Audit, ONSS, and INASTI 
 

“What I particularly value is the direct impact of the work: the analyses produced can influence public debates and policies.”

How does your PhD still support your work today?

The PhD gave me much more than technical skills.
It taught me how to:
•    structure complex problems 
•    work rigorously with data 
•    formulate relevant research questions 
•    manage long-term projects 
Above all, it gave me a working method that is transferable to almost any context.
 

What advice would you give to someone considering a PhD?

First and foremost: be curious and motivated.
A PhD is not reserved for an elite. It is accessible to anyone who enjoys learning, asking questions, and exploring a topic in depth.
However, you need to be ready to:
•    step out of your comfort zone 
•    embrace uncertainty 
•    challenge your own ideas 
 

“It is a demanding experience, but incredibly rewarding both intellectually and personally.”

Looking back, would you make the same choice again?

Without hesitation.
I would not trade this experience for professional experience or financial gain. The PhD gave me a broader perspective and analytical skills that I would not have developed otherwise.
But beyond the skills, what left the strongest impression on me were the people and collaborations.
 

“A PhD is as much a human adventure as it is an academic journey.”