Marco Mancini

Lead Data Scientist

London, England, United Kingdom

About

Lead Data Scientist with 4+ years of professional experience in the areas of machine learning, deep learning and NLP, with an insane passion for getting things done. My main research projects have been around Sequence/Long Sequence Classification, Text/Information Extraction, Language Modelling, Sentiment Classification, Question Answering and Text Normalisation, both in supervised and unsupervised scenarios. From a Cloud engineering perspective I have experience with both AWS and Azure, exploiting Kubernetes as container orchestrator. I am able to build scalable ML pipelines using the most recent frameworks like Argo and Knative, being able to set up and harden the related K8 Manifests. I am mostly a Python coder, but I also have experience with Scala and Play framework in order to build scalable API. I am a polyhedric professional, always hunger for learning and excited to face new challenges and interesting problems. I always had to collaborate across multiple teams and multiple time-zones, and I am used to present my work to both a technical and non-technical audience.

Experience

  • Viaggi at Career Break
    Sep 2025 - Present · 11 mos

  • iManage (4 yrs 9 mos)
    • Lead Data Scientist
      Aug 2022 - Sep 2025 · 3 yrs 2 mos

    • Senior Data Scientist
      Jan 2022 - Dec 2022 · 1 yr

    • Specialist Data Scientist
      Jan 2021 - Dec 2021 · 1 yr

  • VUI, Inc. (2 yrs 10 mos)
    • Research Scientist ||
      May 2020 - Dec 2020 · 8 mos

    • Research Scientist
      Mar 2019 - May 2020 · 1 yr 3 mos

    • Research Intern
      Mar 2018 - Mar 2019 · 1 yr 1 mo

  • Università degli Studi di Trento (1 yr 7 mos)
    • Internship - Signals & Interactive System Lab
      Feb 2018 - Mar 2019 · 1 yr 2 mos

    • Research Project - Machine Learning Lab
      Sep 2017 - Jan 2018 · 5 mos

      The objective of this research was to firstly to learn the Input Convex Neural Network (ICNN) model and its implementation, a Neural Network model proposed by Brandon Amos et Al. in 2017 (https://arxiv.org/abs/1609.07152). Then the model was adapted in order to accomplish a recommendation task, in particular I have implemented an interactive algorithm in order to recommend new food products based on a set of food products the user already likes. The implementation and the related report can be found on my github profile.