Jesus Lago

Science Manager at Amazon

Barcelona, Catalonia, Spain

About

As the head of science and analytics in the EU Prime & Marketing Org, I manage a team of 20 scientists, economists, business intelligence engineers, and data engineers working across a wide range of customer problems. My team works across a large domain of problems: i) Developing Bayesian Geo-RCT models for marketing measurement, b) Building attribution-methods for evergreen marketing measurement, c) Building GenAI-based products to facilitate data analytics by non-technical, d) Recommendation systems and causal inferencing for Prime-exclusive promotions, e) Modeling the business impact of speed changes in the Amazon delivery network, f) Econometric models to model impact of Prime benefits, g) Data analytics for SVP-level business reviews, i) Building the suite of all analytic products of Prime & Marketing in EU (200+ PMs), and supporting them with data-drive business decision. My team has published packages in causal inferencing (e.g. https://github.com/amazon-science/causal-validation), and has authored several scientific papers. I am also the developer of optidef, a Latex library for defining optimization problems, and of the epftoolbox python library, an open-access library for driving research in electricity price forecasting. You can find more about me and my current research in my personal site linked below.

Experience

  • Amazon (5 yrs 7 mos)
    • Applied Science Manager
      Apr 2024 - Present · 2 yrs 4 mos

      As the head of science and analytics in the EU Prime & Marketing Org, I manage a team of 20 scientists, economists, business intelligence engineers, and data engineers working across a wide range of customer problems. My team works across a large domain of problems: i) Developing Bayesian Geo-RCT models for marketing measurement, b) Building attribution-methods for evergreen marketing measurement, c) Building GenAI-based products to facilitate data analytics by non-technical, d) Recommendation systems and causal inferencing for Prime-exclusive promotions, e) Modeling the business impact of speed changes in the Amazon delivery network, f) Econometric models to model impact of Prime benefits, g) Data analytics for SVP-level business reviews, i) Building the suite of all analytic products of Prime & Marketing in EU (200+ PMs), and supporting them with data-drive business decision. My team has published packages in causal inferencing (e.g. https://github.com/amazon-science/causal-validation), and has authored several scientific papers.

    • Applied Science Manager
      Apr 2023 - Apr 2024 · 1 yr 1 mo

      Science manager of the ML team in the EU Prime & Marketing Tech organization. I managed a team of 6 applied scientists and acted as the science lead of organization. My team worked across a large domain of problems: i) measuring customer engagement via hidden Markov models; ii) measuring marketing impact of advertising in video marketing campaigns; iii) evaluating downstream valuation of discounts via causal inference.

    • Senior Applied Scientist
      Aug 2022 - Apr 2023 · 9 mos

      Science Lead of the EU Prime & Marketing Org. Worked across a variety of science problems across a team of scientists, engineers, and BIEs. Projects: Building and designing recommendation algorithms for next action prediction. Building forecasting models for deals recommendations. Measurement of marketing channels via causal inferencing and experimentation.

  • Researcher at VITO
    Sep 2016 - Dec 2020 · 4 yrs 4 mos

    Development of machine learning, data analysis, and optimization algorithms that facilitate the energy transition and the integration of renewable sources.

  • Researcher at EnergyVille
    Sep 2016 - Dec 2020 · 4 yrs 4 mos

    Development of machine learning, data analysis, and optimization algorithms that facilitate the energy transition and the integration of renewable sources.

  • IMTEK, University of Freiburg - Control and Optimization Laboratory ()
    • Research Scientist
      Jun 2014 - Aug 2016 · 2 yrs 3 mos

      - Controlling a tethered kite for energy generation in the field of airborne wind energy (AWE). In particular, implementation of a NMPC scheme in order to track periodic optimal trajectories in real flight conditions. - Modelling and solving periodic optimal control problems to generate flight trajectories that maximize the extracted energy. - Developing a Latex library for defining optimization problems and designing exercises for a book on Numerical Optimal Control. - Development of high frequency readout system for 3 biomedical sensors used for different blood-related measurements.

    • Assistant Lecturer
      May 2014 - Aug 2016 · 2 yrs 4 mos

      - Lecturer assistant in the lecture of Modelling and System Identification. Main tasks: creating and tutoring exercise sessions, creating and grading exams and supervising students. - Tutor in the exercises sessions of the lecture of Micro-mechanics of the Master of Microsystems. Main tasks: solving doubts, guiding students with the course issues and grading exams.