Rafael Alcalde is a seasoned mentor of entrepreneurs in Madrid. He is a computer scientist, but also has an extensive work experience in recruiting and consulting, which gives him a broader perspective in what it takes to start up – including finding the right talent for it. He is now specialized in bots, applying natural language modeling and analytics to serve customers or prospects better, in a semi-automated way. Many entrepreneurs come to him to get insights on the what and how of business models in which data is a key asset. We discussed the initial versions of the DataSET curriculum with him and the overall approach using the Business Canvas, and he shared some of his views with us.

Miguel-Angel: Tell us a bit about you, your current position and your experience in mentoring entrepreneurs.

Rafael: Currently I’m CEO at QrowdMakers, a startup that specializes in developing “bots” (or if you prefer a more formal name, agents for Intelligent Process Automation) for diverse applications and sectors. I’ve been mentoring startups since 2012 at IE Business School and in other incubators and business centers in the vibrant Spanish “entrepreneur scene” that you can find in Madrid.

MA: Have you mentored entrepreneurs in projects that were based on data technologies, e.g. machine learning, Big Data, etc.?

R: Yes, quite a few, for example in domains related to insurance and telco. It seems that entrepreneurs and students value my background in computing and that I am an “IT guy” in a sense, so I get many appointments from potential data-driven or data-focused business.

MA: In your view, which are the main gaps these entrepreneurs have in their skills to effectively propose business models that extract value from data and that are feasible? Which are the topics

R: The main gap is in their lack of skills about concrete data management and handling tasks and technologies. They have typically read a lot about business value of data but almost nothing about the underlying technologies and methods. So, they are not able to create something valuable for the market and they spend and maybe waste money on IT people that is not sufficiently engaged with their project. In consequence, there is a huge problem and their expectations are not fulfilled. The problem essentially lies on a need to get a deeper understanding of the internals. Particularly, some exposure at a high level to the processes and the pipelines needed to create machine learning models is extremely useful to understand possibilities and limitations. This does not mean teaching math or programming, it can be done nowadays with graphical, friendly tools of a complexity similar to that of a GUI-based Business Intelligence tool. Also, understanding the cloud as a computing platform, and the costs associated is required to be realistic especially if business scalability is key.

 MA: Do you believe every entrepreneur should have some training or exposure to the impact, benefits and techniques available to get value from data, even if their business ideas are not related to machine learning or Big Data directly?

R: Sure, it’s a must nowadays. If you are an entrepreneur you should be thinking how to develop a business idea through technology. There is a great opportunity using properly data Science or AI in many domains. We live in a new era and the most scalable and profitable business will come from these technology applications. Entrepreneurs do not need to become data scientist, but they need to understand what data scientists and data engineers do, and even incorporate some part of their jargon.

MA:  How would you think entrepreneurs should be trained in “data skills”? As part of their regular startup training/mentoring process? As a separate course? Do you have some ideas on how to train them and the key topics that should be addressed, e.g. using the canvas, hands-on experiences, business case tutorials?

R: Schools typically do not include this training in their offering, and it will make sense to include it, maybe as a track for those entrepreneurs that really need it for their project, and not for everybody. This will avoid a lot of startup failures I’ve seen in the last 10 years. It’s as important as the Lean Startup methodology is for some projects. So the idea of a concise canvas can be a good starting point, but then there is a need to cover the specifics of businesses based on data, including an account of enabling technologies, the role of different analytical techniques and a practical understanding of how data scientist work routinely and how to constantly evaluate, challenge and revise analytic models.

MA: Do you think that innovation in Europe would benefit from entrepreneurs that are better trained in “data skills”? How do you see the situation particularly in your country?

R: It is indeed. Europe is maybe not the best place for technology startups as we would like it to be. Spain is not different in that, maybe in part because we have a culture of too much risk avoidance. But the future of any country depends on some people risking their time and work and eventually creating disruptive innovations or discoveries. Those innovations may become the engine for the next hundred years, as happened with sailors and discoverers in the XV-XVII century in Spain or in US in the 40’s to 70s, investing a lot of money in R&D and supporting the activities of entrepreneurs. Data skills can be a fertile ground for such innovations, and developing human capacities is the only way to achieve them.

Interviewed by: Miguel-Ángel Sicilia Urbán, The University of Alcala