“I have a doctorate in engineering, which means I can discuss in-depth technical topics. However, I feel that I am a people-oriented technology leader and I communicate fluently in a clear and understandable way to different audiences. ”

Timo studied statistical mathematics, physics and signal processing and worked as a researcher in the computational systems biology group in Tampere, before moving to Amazon.com to develop demand forecasting models. Under the title forecasting research scientist Timo developed demand forecasting models for new products and discount sales, among other things. More specifically, the topic was the integration of multiple data sources into demand forecasting and the consideration of uncertainty (integrative and probabilistic demand forecasting).

“Demand for e-commerce is affected by many internal and external factors: price, product presentation, marketing channel, availability, competitor situation, season, etc. Producing forecasts requires a working data platform on which to build a machine learning model that learns the impact of different inputs on demand for each product. This (unbiased) demand forecast can be used, among other things, in warehousing and logistics planning, as well as in optimizing sales and prices. A probability-based forecast can take into account the inaccuracy of the forecast, which has an impact on decision-making.

Ideally, forecasts are fed into an automated sourcing system that controls inventory levels. In practice, there is also a need for a way for people to participate in decision-making if automatic predictions are not accurate enough. ”

In 2014, Timo became Quva Oy's Chief Technology Officer and was the architect and one of the main developers of the Quva Flow quality control system. During that time, both industry standard methods (lean six sigma) and more modern methods became familiar.

Timo's own areas are software architecture, cloud architecture, technical development plans and project management of implementation using Agile and Scrum methods. For those interested in technology, Timo's toolkit includes AWS, GCP, Azure, Python, Scala, Javascript, TypeScript, C / C ++, and Databricks / Spark, React, Redux, Theano, Tensorflow, Keras, Pandas, Numpy and Scikit-Learn.