Explore the 2023 case studies

Have a look at the case studies of this year's participating companies.


Shell 01

Concerns about machine learning (ML) applications' environmental impact, especially their carbon footprint, have been raised as they become more commonplace in a variety of industries. ML models have high computational requirements, and the power-hungry hardware they need can result in substantial greenhouse gas emissions, which contribute to climate change. As a result, there is a growing need to understand and mitigate the carbon footprint of machine learning, both at the individual model and system level, to ensure that their benefits do not come at a prohibitive cost to the environment. In this project, we aim to measure the carbon footprint of large language models and propose strategies to reduce it, including shifting time and location for training, adopting more energy-efficient hardware, and designing sustainable training and deployment practices. By doing so, we hope to contribute to the development of more environmentally responsible AI technologies.

The main challenge to our students will be to greatly improve a prototype application to measure energy and carbon intensities of machine learning models, in particular large language models such as GPT. Fundamental goals of the application development include linkage of existing prototype to carbon monitoring solutions along with packaging of the code for easy re-use by data scientists at Shell, with potential for your work to be open sourced for use beyond Shell.
Additionally, you will perform experiments with various ML models to better understand the energy and carbon impacts of various setups. Delivery of a report of your experiments, findings and recommendations will be an end goal of this project.

Research questions
Can we measure the carbon footprint of training our machine learning models?
Can we help reduce the carbon footprint by choosing the time and the location for our computation where the energy used comes mostly from renewable sources?

What we expect of you
Shell ETCA, Amsterdam (≥ 1 day / week) Virtual (rest of week). Align days within team. Commitment:  ≥ 4 days per week
DS/AI/technical students with experience in DS/AI and interest in reducing carbon intensities of AI.
Requirement Skills: Python, Experience with 1 or more machine learning libraries (e.g. Tensorflow), comfortable with command line tools in UNIX based systems.
Desirable Skills: Time series analysis, data visualization, software engineering best practices (e.g.unit testing, code management in tools like Git, Agile development), cloud platform (e.g. Azure, AWS)
Permission to work in NL is required
Internship compensation and travel expenses provided.

Shell 02

Shell is growing exponentially in the E-Mobility space! As you may already know, Shell has been committed to expand its electric vehicle (EV) charging network from 60,000 now to 500,000 chargers in 2025 across the world. This is exactly why we want to motivate you to come work with us on this year’s challenge that will help Shell develop more intelligent and efficient way of operating our EV charging sites.

For this case, we zoom to an EV charging site where a 300 kW/360 kWh stationary battery is deployed to support EV charging operations. This site has two DC chargers, and the stationary battery is an integral part of the local charging system. Inclusion of batteries in overall workflow introduces unique challenges as well as opportunities. While there is a cost associated with every battery charged and discharge operation, it can also be used for temporal arbitrage. Your challenge is then to develop an AI-driven solution to maximize both profit and battery lifespan depending on several factors, such as charging demand, electricity price and battery characteristics.

What we ask of you
Full commitment to the competition and minimal 1 day per week (on the agreed day) to our site at Shell’s Energy Transition Campus Amsterdam.
• Students in STEM related field with experience in ML/AI and good programming and communication skills.
• Compensation provided and travel expenses are reimbursed if you are not in possession of student OV card.


Make KPN’s telecom network future proof by using predictive maintenance techniques
KPN has been connecting Dutch society for more than 100 years. We started with postal mail and
phone calls. And now everyone is always connected to the internet. Over the years we have
undergone big technological developments and we have set up an exceptional network. KPN’s
network is everywhere around you, inside and outside your home. Sometimes you can see parts of
this network in your own neighbourhood such as street cabinets and mobile masts, the modem in
your home, but most of our network is underground.

The challenge
Making sure that the network is operating at peak performance is a big undertaking. Equipment in
the network, in home and underground, can break down due to numerous causes. If we can predict
when equipment in the network is likely to break or when issues in home will occur, we can
anticipate on that and adjust our maintenance planning or prevent impact.
By applying predictive maintenance, you can help ensure that KPN’s network is operating at peak
performance and help to further decrease our emission footprint and stay on top in the list of most
sustainable telecom operations in the world.

Case 1: Predict network element failures
Analyse telecom network related data, such as routes through the network and element
performance data like traffic and temperature to predict network element failures
Collaborate with other team members data scientists from KPN and business stakeholders to
understand, assess, and implement the predictive maintenance solution
Develop a predictive maintenance model using machine learning or other analytical
Set up an implementation plan

Case 2: Predict TV issues
• Analyse customer TV measurement data, along with any other relevant data sources to
detect/predict TV issues for the customers and find solutions to resolve those problems and
improve customer experience.
• Collaborate with other team members data scientists from KPN and business stakeholders to
understand, assess and implement the predictive maintenance solution
• Develop a predictive maintenance model using machine learning or other analytical
• Set up an implementation plan

What we ask of you?
• Full commitment to the competition 4 days per week (non-weekend) and minimal 1 day per
week (on the agreed day) to our site at KPN Amsterdam or Amersfoort.
• For this internship you need a strong background in data analysis and predictive modelling.
• Ideally, you have experience in Python programming using the Pandas package.
• Preferred master student
• For this internship we ask you to request a certificate of conduct (VOG) as part of the
onboarding process.

What does KPN offer?
• This case will give you the opportunity to work on a real-world project at the heart of
analytics at KPN with 45 experienced data scientist colleagues.
• Work together in the office on Monday in Amersfoort, near the Amersfoort CS station and
Thursday in Amsterdam, near the Amsterdam Sloterdijk station
• A market rate internship compensation
• Travel expenses provided if you are not in possession of student OV card.
• A laptop to work on

We are looking forward to hearing from you!

Tata Steel 01

The reduction of energy consumption in steelmaking is one of the most significant challenges nowadays. Ensuring the industry's competitiveness and minimizing environmental impacts, such as CO2 emissions, is crucial. Tata Steel in the Netherlands (TSN) committed to reducing CO2 emissions by 35 - 40% by 2030 and being CO2-neutral by 2045. Therefore, projects to achieve more efficient production, monitor the processes and increase their quality will lead to significant sustainable steelmaking and assist TSN in reaching its emission targets.
Steel production is a very complex process, and applying advanced digital tools is essential for optimizing the entire production efficiency. A huge potential for energy efficiency optimization and CO2 emission reduction is in the control of the temperature of hot metal. Hot metal (or pig iron) is an intermediate product in steel production. Hot metal is produced at the blast furnaces and transported in so-called torpedo railcars to the oxygen steel plant where it is poured into hot metal ladles. The thermal logistics management of the torpedo railcars and hot metal ladles is a complex interaction of several processes.

Our challenge to you: can you develop an application to support real-time decision-making on the thermal logistics management of the railcars or hot metal ladles?
The goal is to create a logistic optimization algorithm from the current straightforward methodology (first-in first-out) towards a more sophisticated one, such as reinforcement learning. You will have access to annotated process data and a pre-developed digital twin. The digital twin is developed from physical models to calculate the real-time thermal state of each railcar or ladle in operation. In addition, the user interface is essential for field application to assist the operator's decisions, and this way, it enables improving energy efficiency in iron & steelmaking processes.

Your summer of AI at Tata Steel
During the summer, you’ll be coached by Tata Steel advanced analytics coaches and gain experience in our agile way of working. You'll have contact with process matter experts to validate preliminary results. And you’ll have the opportunity to be at the heart of the steel industry’s single largest transformation in history and be able to have a positive impact by contributing to TSN’s journey of becoming a clean, green and circular steel company. At the end of the summer, you have learned how to create a proof of concept and work with large scale data systems and digital twins, such that at the end of the summer, we can start implementing this application.

What we ask of you
4 days per week (non-weekend) of which min. 1 day per week on our site at the Tata Steel IJmuiden.

• Preferred master student in STEM related field with experience with ML/AI.

• Compensation provided and travel expenses are reimbursed if you are not in possession of student OV card.

Tata Steel 02

Steel surface quality is one of the most important product properties to our customers (particularly our automotive customers). Hence, there are camera inspection systems placed in the production lines of Tata Steel’s that continuously scan the steel surface for defects. These systems are used to detect and classify surface defects and feed the systems that determine whether the steel meets the quality standards requested by our customers. In some cases, the coil needs to be degraded – in other cases part of the coil needs to be cut out.
Based on experience we know that the defect classification routines built-in the software provided with the camera hardware is not very accurate. The last couple of years, we have successfully applied deep-learning methods to improve this classification. However, we face a number of bottlenecks to truly scale this up:
1. Labelling: we require typically >10k images which takes a significant amount of time from our surface quality experts.
2. Accuracy: despite having labelled >10k images, the model precision typically reaches a maximum plateau around 95%.
3. Parsimony: we currently use a two-step modelling framework (CNN and subsequently a Random Forest) to combine defect image data with tabular data of this image (position on the strip). This is inefficient when it comes to model maintenance, and performance-wise probably suboptimal.
4. Re-useability: camera hardware, settings and steel appearance hamper the re-usability of the developed deep learning models. This greatly increases the required labelling efforts.
On all these aspects, we have ideas to improve our current way of working. For instance, we are very curious whether object detection instead of image classification can reduce the labelling burden. Also the use of unsupervised learning techniques in the labelling process might drastically reduce the labelling effort. The accuracy of our models might be increased by better accommodating for the considerable class imbalance. To reduce the number of models to maintain, we think about a solution in which auxiliary variables are somehow included in the convolutional neural network. On the re-usability, we have been hypothesizing that Generative Adversarial Network (specifically cycle-GANs) might be a way to re-use labelled images measured under different conditions.

Our challenge to you: can you develop a Proof-of-concept on one or more of these topics? We deliberately present a wide range of topics because we want to explore together with the students on which topics they think there is the most to gain. And there is certainly not a lack of data on which to test and trial new methods: we have a cloud environment packed with >100k images at your disposal! By helping us improve steel defect monitoring, you can directly contribute to Tata Steel’s overall quality performance!

Your summer of AI at Tata Steel
During the summer, you’ll be coached by Tata Steel advanced analytics coaches and gain experience in our agile way of working. You'll have contact with process matter experts to validate preliminary results.
At the end of the summer, you have learned how to create a proof of concept and work with large scale data systems, such that at the end of the summer, we can start implementing this system.

What we ask of you
4 days per week (non-weekend) of which min. 1 day per week on our site at the Tata Steel IJmuiden.

• Preferred master student in STEM related field with experience with ML/AI.

• Compensation provided and travel expenses are reimbursed if you are not in possession of student OV card.


”Banking for better, for generations to come” is the purpose of ABN AMRO Bank. We are committed to promoting sustainable business practices and supporting our business clients in achieving their sustainability goals. Our use case this year is an innovative initiative that generates insights to help our business clients become more sustainable.

It's crucial for businesses to adopt sustainable practices that ensure long-term success; however, many businesses struggle to assess the related risks and opportunities accurately. With sustainability becoming an increasingly important factor in business decision-making, we aim to provide valuable insights into areas such as carbon emissions, transitions to green energy, and performance compared to peers. From sustainable financing to green bonds and impact investing, we can help businesses finance their sustainability initiatives.

Your summer of AI at ABN AMRO 
You will join the AI team of the Strategy and Innovation Department and gain practical experience in using AI and open data to solve real-world challenges. In this project, we aim to extract actionable insights from different unstructured data sources with NLP and/or Computer Vision, such as satellite images, annual report, etc. At the end of the summer, you will learn how to tackle a problem head on and deliver a working proof-of-concept to a team of experts in sustainability.

What we ask of you
4 days per week (non-weekend) of which min. 1 day per week in our office at the Zuidas Amsterdam
•Preferred master student in STEM related field with some experience in programming (Python) and ML/AI
• Experience with unstructured data, such as NLP, computer vision is a plus
• Provide a VOG at the start of the program (
• Internship compensation and travel expenses provided.

Sustainability; AI for good; open data; python; unstructured data; computer vision; natural language processing

As a part of the FD Mediagroep, holds the position of the market leader in providing Dutch business information and insights. Our clients rely on us for qualitative, up-to-date, and in-depth insights into the latest company news. We cater to the needs of every business professional who wishes to establish robust business relationships and boost their sales. Our data is accessible via various channels, including an online portal, APIs for system integration, online dashboards for analysis, and data delivery. By providing these solutions, we aim to help professionals achieve more success in their business endeavors.

We want to link companies' websites to legal entities from Kamer van Koophandel (KVK) and to the buildings in Basisregistratie Adressen en Gebouwen (BAG). An important piece of information we use for this are addresses. We found that currently there’s no Deep Learning model that is good at extracting Dutch addresses from text. We want to train a model that extracts Dutch addresses well. At we have data of millions of Dutch companies and their websites, including large amounts of hand-labelled data. For this case, we have data consisting of pairs of websites and addresses that we extracted using regular expressions. Training the first effective Dutch address extractor would already be quite a feat, but if time allows there are different options for expanding upon that, depending on the interests and creativity of the team.

Your summer of AI at 
You will work closely with our data science team of four and have dedicated supervisors. You will gain practical experience in applying state-of-the-art Deep Learning / NLP techniques for solving real-world problems as a Scrum team. You will also enjoy the friendly atmosphere of and are welcome to join our weekly drinks.

What we ask of you
A solid foundation of Python knowledge
Experience with Deep Learning, NLP and git is a plus
Enthusiasm for teamwork
We expect you to be at our Amsterdam office 2-3 days a week in order to meet your team face-to-face

What we offer
Market rate internship compensation
Travel expenses are covered
A laptop to work on
An office that’s easy to reach from Amstel Station
Supportive and fun work environment in a mid-sized company
Enthusiastic and helpful colleagues around you

South Pole

Introduction to South Pole:
South Pole is an energetic, global company offering comprehensive sustainability solutions and services. With offices spanning all continents across the globe, we strive to create a sustainable society and economy that positively impacts our climate, ecosystems and developing communities. With our solutions, we inspire and enable our customers to create value from sustainability-related activities. Our Data and Insight team help South Pole design and execute the most ambitious and scalable climate impact initiatives whilst reducing the administrative effort to zero.

Purpose of Project:
The purpose of the project is to support our ongoing GHG accounting efforts by working directly with our EIA experts to develop an NLP based predictive model to identify the best match emission factor to apply according to client data. This will involve working with our data science team for model build and our data platform team to also support the build of an API endpoint.

GHG accounting is the product with which we calculate the emissions of clients (from small to large companies) and thereafter can propose mitigation strategies. For this we gather, create, analyse and store detailed emission factors across scope 1, 2 and 3 (‘activities’). An emission factor is a coefficient that describes the rate at which a given activity releases greenhouse gases (GHG) into the atmosphere.

Your Summer of AI at South Pole:
A team of experts from South Pole are already working on this and you will be embedded into this project team alongside a data science mentor. You will work on a “real life” data science task, meaning you will also need to assist in getting the data for your training set. You will have full access to our GHG experts and data team members to support you with this.
At the end of the summer you will have gained:
a solid understanding of GHG accounting, emission factors and mitigation strategies
produced a predictive model using NLP and AI techniques
contributed to the build of an API
gained an understanding of what it means to build digital data products
You will work on our infrastructure using Python as the primary coding language, supported by our MLOps infrastructure and GCP

The assignment:
Be in our office in the Zuidas at least 2-3 days per week
Master’s degree student with solid experience in Python and advanced data science. Software Engineering experience is beneficial.
Internship compensation is provided as per market rate
Allowed to work and live in the Netherlands without sponsorship

We look forward to welcoming you to South Pole!

Kickstart AI in partnership with the Zero Hunger Lab

Kickstart AI was founded by four Dutch icons with big orange hearts who believe that AI will shape our collective future 🧡: ING, KLM, NS and Ahold Delhaize. The goal? Accelerate the adoption of AI in the Netherlands, with a focus on building a collective future where AI is a positive force for good, powered by collaboration, innovation and knowledge sharing. As such, we are the coalition of ‘the doing’. We work with our partners, academics and experts to develop AI solutions for societal challenges, building a blossoming AI community along the way.
As a coalition of ‘the doing’, we work in partnership with a number of universities to accelerate the adoption of AI through applied methodologies, including Tilburg University's Zero Hunger Lab.
The Zero Hunger Lab focuses on addressing food insecurity issues by applying data science methodologies. Our team of multidisciplinary researchers works to identify and solve challenges related to food insecurity. For instance, we optimize food baskets using operations research for the World Food Programme, enhance supply chain management with data analytics, and forecast demand through machine learning to minimize food waste.
A prevalent issue in many food-insecure nations is the limited availability of relevant data. Information on factors such as food prices, crop yields, conflict data, migration data, and current levels of food insecurity might be missing or infrequently updated. This data is crucial for forecasting food insecurity since high-quality predictions could mobilize humanitarian resources, like food packets, in advance. Shifting from a reactive to a proactive aid policy could save millions of lives in regions such as the Horn of Africa.

The Challenge
We are interested in improving the data processing surrounding food insecurity and predicting food insecurity ahead of time. The standard system to classify food-insecure regions is the IPC (Integrated Food Security Phase Classification), which categorizes food insecurity into five phases, ranging from the least to the most insecure (see Figure 1 for an example).

Figure 1: Example of an IPC classification in the Horn Of Africa.
Your task is to uncover interesting patterns to explain or predict these food insecurity phases in regions in the Horn of Africa. We will provide you with datasets from classical sources, including monthly climate and weather data, food prices from the WFP, and conflict data from an academic database. Additionally, we will supply standardized data from common news sources. Prior research has shown that NLP methods applied to news sources can predict food insecurity in advance [BSF23], and we are curious to see if you can expand upon this. Lastly, we encourage you to develop new and innovative ideas to track or predict food insecurity using public data we haven’t considered before, such as data from satellite images.

What We Ask of You

A background in data science/machine learning or other technical domains (e.g., master students of AI, Econometrics, Data Science, Computer Science, etc.), or strong data science skills. Programming knowledge is required, such as working with Python libraries like Pandas and Sklearn or equivalents in other languages.
Some background knowledge in NLP (classical methods or Deep Learning).
An interest in a project that considers context, rather than being solely technical. For example, you might want to learn about the recent history of the Horn of Africa to better understand the situation.
Experience with applying data science methods on real-world data. For instance, working with sub-optimal data containing missing values and requiring data processing. As a scientific lab, we are not seeking a perfect solution, but rather interesting insights!

What We Offer
A fun, startup-like environment. You will work at our brand new Delft Kickstart AI office a minimum of one day per week, although you are welcome everyday of the week. We expect you to be available four days per week to work on this project. Our office is complete with a playstation, table tennis tables, free coffee and tea and of course, weekly borrels!
The opportunity to become part of Kickstart AI’s data science community, with specific regular engagement with our four partner companies: ING, Ahold Delhaize, NS and KLM. You will receive invitations to attend Kickstart AI’s exclusive events, speak with Heads of AI from our partner companies and engage regularly with their data scientists and engineers.
Work with both corporate companies, and universities, on a project that is focused on applied AI, for good. In joining our efforts, you will not only expand your skill set and knowledge, but also contribute to a meaningful mission!

[BSF23] Ananth Balashankar, Lakshminarayanan Subramanian, and Samuel P. Fraiberger. Predicting food crises using news streams. Science Advances, 9(9):eabm3449, 2023.

Nederlandse Spoorwegen

NS strives to provide convenient, fast, safe, affordable, and sustainable travel. NS is developing into a broad service provider that works with other companies to offer the customer the option of 'smart' travel based on the most up-to-date information. Convenient travel, from door to door. [1]
OV-fiets is the readily accessible rental bicycle for the last part of your trip. If you arrive at the station by train, you can quickly and easily rent a public transport bicycle and cycle on to your appointment, go visit friends and family, go to the museum or attend a business lunch. There are almost 300 rental locations: at many stations, at bus or metro stops and at P+R sites. [2]
Currently, NS cannot provide information about the short-term availability of this mobility service, due to which customers are limited in being able to reliably plan their door-to-door journeys. As the first step in addressing this problem, NS is currently developing a data product to provide information on OV-fiets availability within the upcoming 48 hours.

Research questions
How can/should OV-fiets availability be defined, measured and forecast?
What is the best way to communicate OV-fiets availability forecasts?
What is the best way to measure the impact of OV-fiets availability forecasts?

What you will do
You will research and develop a model to forecast OV-fiets availability at an OV-fiets rental location, and compare its predictive properties to the existing baseline models. This will involve for example field work, user experience research, and applying predictive modelling techniques.

Where will you work
    At the NS's head-office located right next to Utrecht Central train station
    Among and with an experienced group of NS Data Scientists and Engineers
    In close collaboration with NS domain experts
    On NS's Microsoft Azure platform, including:
        Azure Databricks environment
        Sandbox Azure development subscription
        DevOps services, including Boards, Repos, and Pipelines
    Communications through NS Slack service

What we expect from you

Commitment to availability: at least 7 out of 8 weeks
Shared values: team play, customer-focus, open-mindedness, transparency
Physical attendance: meet in the office at least one day per week
Digital attendance: check in online on a daily basis

[1] NS mission
[2] OV-fiets frequently asked questions


Problem statement
Our sales team regularly meets with customers and records the details of each visit in the CRM system. We want to analyse these reports to identify common customer needs and preferences, so that we can better tailor our sales approach and improve our sales effectiveness.

Data collection and pre-processing
We collect customer visit reports from the CRM system and pre-process the data to remove irrelevant information such as timestamps, sales rep names, and other identifying information. We also clean the text data by removing punctuation, stop words, and performing stemming or lemmatization as necessary.

Data analysis
We use NLP techniques to analyse the text data and extract key insights. Here are some possible analysis techniques:
Topic modelling: We can use topic modelling techniques such as Latent Dirichlet Allocation (LDA) to identify common themes or topics that appear in the customer visit reports. This can help us understand what customers are most interested in and what their top pain points are.
Sentiment analysis: We can use sentiment analysis techniques to identify whether customers are expressing positive, negative, or neutral sentiment in their visit reports. This can help us understand how customers feel about our products or services and what we can do to improve their experience.
Named entity recognition: We can use named entity recognition techniques to identify specific products, features, or services that customers are mentioning in their visit reports. This can help us identify which products or features are most important to customers and what we should focus on in our sales pitch.

Insights and action
Once we have analysed the customer visit reports, we can use the insights to improve our sales effectiveness. Here are some possible actions we could take:
Tailor our sales pitch: Based on the common themes and topics identified through topic modelling, we can tailor our sales pitch to better address customer needs and pain points. For example, if customers frequently mention a need for better customer support, we can emphasize our strong customer service in our sales pitch.
Improve our products or services: Based on the sentiment analysis results, we can identify areas where customers are expressing negative sentiment and take steps to improve those aspects of our products or services. For example, if customers are frequently complaining about long wait times on the phone, we can improve our call centre processes to reduce wait times.
Focus our sales efforts: Based on the named entity recognition results, we can identify which products or features are most important to customers and focus our sales efforts on those areas. For example, if customers are frequently mentioning a particular product feature, we can make sure our sales reps are trained to highlight that feature in their sales pitch.

Stay up to date!

Subscribe to the newsletter to stay informed about the Summer of AI and make sure that you are among the first when the challenges of the companies go public.

We'll only use your e-mail to send you updates when challenges become available!
* We don't share your data. See our Privacy Policy
You've subscribed!
Oops! Something went wrong while submitting the form.

Latest news