We make Wolt even more intelligent
We empower Wolt employees to make smarter, data-informed decisions.
We champion how we use, collect, model and present data — and make the complex world of data simple for our teams.
Our data teams
Data science at Wolt
Data Scientists at Wolt are working with machine learning for the most important processes and parts of Wolt’s online delivery platform and business operations that affect all the 23 countries we operate in.
Embedded in cross-functional business and product teams, we help teams to solve business problems with anything from forecasting to predicting and optimizing processes. Data scientists have an end-to-end role in the team. They help investigate the business problem, build models, take them to production, and maintain them.
Our data scientists work on things like:
Logistics Optimization: We improve our logistics platform to make Wolt’s services as efficient as possible – and we do this by being smart about how we allocate tasks to available couriers. We estimate several different time steps during the delivery journey, and use those as inputs into our optimization process.
Delivery ETA: Our models produce around 57 billion delivery time estimates daily, and those are used in the different pages of our app to heavily influence the overall customer experience.
Personalization: We provide personalization and recommendations at multiple places along the customer journey. Making recommendations is complex — changes in the customer’s location also changes the selection of venues available to them, while user intent can vary a lot. We generate recommendations for over seven million users daily.
Marketing: We’re building up our marketing data platform that helps us automate campaigns and optimize spend across channels while understanding the impact of our marketing activities.
Product Analytics team
Our product data analysts are embedded in our product teams to be close to the decisions we’re making for our product. We play a key role in helping teams understand our customers and their behavior.
We help our teams focus on things that have the biggest impact – always ensuring we make data-informed decisions relating to our product. We tell stories with data, translating numbers into actionable insights for our teams.
Our Product Analysts work on things like:
Experimentation of new ideas — we use our internal A/B testing platform to validate and measure the impact of any changes we make. Understanding that is key to how we build Wolt.
Research projects to deep-dive into user data patterns. For example, investigating the main drivers of customer churn to help us better retain customers.
Impact analysis to quantify the impact of potential changes. For instance, understanding the effect on purchase volume if the delivery radius of a city is reduced for some reason. This helps us shed light on the most impactful problems to fix, and also which opportunities to go for.
BI Platform team
Wolt invests heavily in self-service analytics. We enable our colleagues at Wolt to make smarter, data-informed decisions by providing data and the tools to access and analyze it. With a global team of thousands of employees, we do this at scale. Being able to make a small core team scale to serve thousands of data users has been one of our true success stories. We do this by making our data as user friendly and well-documented as possible, and providing self-study materials and training sessions.
Naturally, self-service analytics only works if the data is modeled in an user-friendly way, while maintaining data warehouse and metadata with great care. This is something our team excels in.
Our BI Platform team works on things like:
Our centrally maintained Looker data integrations, data pipelines, data models and dashboards
Writing complex SQL queries
Looking into how we can empower our teams to use data
Training our teams to work with data in an effective and independent manner
Providing user-friendly documentation and tutorials for everyone at Wolt working with data
Creating best practices on how to develop data and utilize it in daily work
Providing hands on help and support to our users on how to use and interpret our data
How we work
Our vision — the whole city, delivered
Our vision is big — the whole city, delivered within minutes. To get there, we’re connecting three parties: our customers, courier partners and merchants. Seamlessly and in real-time. This three-sided marketplace calls for a balance that has to be just right. Our business is about constant optimisation of all these factors. And it means we need data about everything that might impact running this platform.
We need data not only about what happens in our Wolt platform but also from what happens in the real world — data about things like weather or public holidays— whatever might affect our operations. Typically the data we use comes in the form of geospatial and temporal data and we use it to dive deep into understanding our customers’, courier partners’ and merchants’ behavior to optimize our platform.
Data at scale
We operate in 23 countries and 400+ cities. Our products are used by millions of users, tens of thousands of courier partners and thousands of merchants. We have a lot of opportunities for automation to optimize our processes for the volumes.
Inside Wolt, we have thousands of colleagues actively using data in their work. We’re all about empowering our people to use data in the best possible way, at scale.
Most of our data is collected and processed in real-time. For instance, many of our use cases for machine learning require us to have real-time information about what’s going on in our platform right now.
Data is key to delivering business value
We build scalable and robust solutions, follow operational and business metrics, draw insights and offer a delightful user experience for all data end-users.
We’re all about finding the balance between business value and tech. Beyond data tooling, we focus on moving the needle to get Wolt towards our mission – and we do this by delivering business value and insights.
We work with a variety of technologies
Our data tech stack covers but isn’t limited: Python, SQL, AWS, Looker, Snowflake, Airflow, Kafka, Spark, Docker, K8s, Jenkins.
We have a dedicated stack for machine learning, where we rely on technologies like Flyte, MLflow and Seldon Core, with Kubernetes compute clusters.
Were built on ownership and empowerment
We value autonomy and ownership — we’re trusted as the experts of data to do the right thing. We’re kind and friendly, but also not afraid to question the requests we get or ideas that come up as we always prioritize based on impact.
We work closely with the teams we support — whether as embedded members or as those who empower teams to use data. We respect each other’s ideas, opinions and expertise. Together we aim high, but stay humble and help each other out.
Hear it from the team
"Everyone is friendly here, there is no place for big egos. When we have different opinions, we talk about them openly, discuss facts and come up with the best solution for the problem at hand."
— Juan, Data Scientist
"Being embedded in product teams, I get to learn from developers, data scientists, and product leads. It’s valuable to get different perspectives to improve my own work."
— Vera, Product Data Analyst
"I learn something new every day since we use top-notch tools, stacks, and systems. One of the best things about Wolt, is that it has excellent developers and bigger teams, so you get to see what all the other engineers do, as well as see different aspects of data."
— Valeri, Business Intelligence Analyst
