Identify data-related business opportunities

Artificial Intelligence, machine learning, and Big Data won’t add much value to your business unless you carefully plan for it. By now, we’ve seen many organisations jumping into data (science) head first, and failing to provide immediate value to their business.


All these new technologies promise to fundamentally change businesses, and being late to the game might be fatal. But without a clear strategy and value proposition, no algorithm will magically add value to your company.


With years of experience in data-driven companies, we help you select and specify projects that immediately add value and impress your stakeholders. We help you to develop a roadmap that starts with simple solutions as a proof of concept, assist you in taking incremental steps towards a more advanced solution, and finally help you transition into a fully productionized solution.


We help you identify opportunities by:

  1. Conduct interviews to understand business goals & data

  2. Design metrics to make goals measurable

  3. Work with your team to generate initial hypothesis for improvements

  4. Run exploratory data analysis to identify, validate, and quantify opportunities

  5. Provide stack-ranked list of options by impact & effort


Implement data-driven decision making

What is the common denominator behind the big tech companies’ success? They all practice data-driven decision making. Meaning they make decisions based on scientific principles to guard themselves against biases like, e.g. confirmation bias, base-rate fallacy, and more. A data-driven approach ensures high-quality decisions in today's fast changing world.


Our workshops will help your leadership team to become data-driven and learn quickly learn about the value of data, avoiding common pitfalls, and eventually consistently make world-class decisions based on objective information.


Data-driven decision making usually operates within a well-defined framework:

  • Define objectives

  • Formulate (testable) hypotheses on how to achieve your objectives

  • Test your hypotheses through analytics, experimentation, and structural models

  • Refine & iterate


Implementing this process, however, is only the first step. It’s most critical to draw the correct conclusions from your data. That’s where real-world data-driven decisions are often challenging. Instead of robust scientific methods we see semi-scientific regressions coerced to validate someone’s intuition. We help your organization with fast but accurate scientific analyses to support robust decisions.


Key methods for unbiased decisions are:

  • Analytics

  • Experimentation

  • Structural Modeling

  • Forecasting

  • Optimization



Analytics is the key ingredient in a data-driven approach. Slicing & dicing your data quickly reveals important insights and sets the foundation for further exploration. A key ingredient to world-class analytics is to deeply understand the difference between correlation and causation. That two quantities move together doesn’t mean they one causes the other.


Ingredient #2 is presentation. Insights need to be quickly digestible, your senior leadership team simply doesn’t have the time to decipher a graph for 15 mins. We’re experts in storytelling with data and ensure any insight is easily digestible.



Experimentation is the common denominator behind the success of all data-driven organizations. It enables them to precisely identify which of their decisions impacts their bottom line, and disentangles causation from correlation.


Together we’ve run thousands of experiments, and implement & refined product experimentation in many large organizations. We’re happy to support you in your journey from correlation to causation.


Structural Modeling

Experimentation is always preferable, but in some situations it’s simply not feasible. That’s where structural modeling comes in. It’s also an area where machine learning cannot help since it’s simply not designed to uncover causal relationships. The fields of statistics and econometrics develop a large set of methods designed precisely to answer these types of questions.


Our team is full of industry experts as well as professors who deeply understand and frequently apply these techniques. 


Build data products

Data, artificial intelligence, and machine learning is on its way to fundamentally change many industries. They allow services that were previously unthinkable, either technically or economically. Top-of-mind examples are self-driving cars, chat bots, and self-landing rockets.


But these are just the shiny revolutions on the surface. The real revolution is happening right below the surface. Google, Facebook, and Amazon’s ranking algorithms are changing how we interact with content and make decisions. Even classical businesses like delivery now heavily rely on machine predictions to optimize routes, order packaging, and traffic patterns.


Many companies understand the opportunity and are aiming to implement these new technologies. Yet few succeed: 87% of machine learning projects fail to be implemented. Key drivers are the technological complexity but even more important is immediate value to the business. Most AI/ML projects fail to deliver any value for the business within a reasonable time frame. 


We leverage our expertise from countless AI/ML projects to help you create immediate value for your business by starting with a simple prototype and only moving towards more advanced technologies after value is proven. As in product development, we start with a minimum viable product that shows e.g. that a certain feature is feasible in production. Then we assess together the potential to add value and only then start using cutting edge technologies. This approach helps us to focus on value, impact, and speed.


From there on we iterate and help you build world-class technology based on years of industry experience and the newest research papers. Leading academic researchers on our team ensure we’re always on the cutting edge of technology.


Our process in short:

  • Create deep understanding of the business problem

  • Build a minimum viable product (MVP) to prove value

  • Iterate towards a  cutting edge solution