Menu

Florent RENUCCI

Paris

En résumé

(I never check my Viadeo account, please contact me through LinkedIn instead).

After three years of intensive university­‐level preparation for entrance exams to top French engineering schools (“Classe Préparatoire aux Grandes Ecoles”) studying mathematics and computer sciences, I specialized in Applied Mathematics, especially Machine Learning and Artificial Intelligence, applied to business problematics, to become a Data Scientist.

Those questions are at the intersection of :

- applied mathematics at a very high level, because the breakthrough heuristics, that allow to outperform the market, lie in the academical research universe.

- computer sciences, to make the link between mathematics and business real.

- the sciences of business (eg. strategy, corporate finance, marketing), to keep in mind the business feasability of the solution along the research project.



This paradigm has guided my educational and professional choices.

My experience strategic consulting and the education I have received from the ESSEC Business School during my MBA gave me a good overview of the sciences of business, and how to take highly valuable data-driven decisions.

The research projects I have worked on, with INSEAD in Paris or with Amadeus North America in Boston, and the Master's degree in research in applied mathematics from the Ecole Normale Superieure, allowed me to work with and learn from very high-skilled researchers in the field of Machine Learning, as well as revealing the underlying patterns behind the data.

And of course, the french engineering school Ecole Centrale Paris taught me to implement, build the bridge between abstract mathematics and concrete business, and realizing ideas - making ideas real.



My goal as a Data Scientist is to use and take advantage of those three visions to solve Big Data problematics and extract intelligence from data (Business Intelligence), basically :

- mathematics : from idea to concept,
- engineering : from concept to code,
- and business : from code to value creation.


Mes compétences :
Langage c
Langage C++
Langage de balisage maitrisé HTML
Objective c
CSS
Javascript
JAVA
Python
VBA
R

Entreprises

  • Goldman Sachs - Data Scientist

    Paris 2014 - maintenant Strategist - People Analytics.

  • Amadeus North America - Data Scientist

    2013 - 2014 - Objectives :

    Building a popularity predictive model of travel searches on airlines websites or on aggregators. In two words, I have to predict today how many people will search for any given flight tomorrow.
    The aim is both to better know the future demand and the overall market to optimize the frequency of the update of the prices and to reduce costs for calculations.

    - Framework :

    I work in the SSP team (Search, Shopping and Pricing) of twenty people in the research center of Amadeus North America in Boston, Massachusetts. I build this model alone, basing on Machine Learning metaheuristics applied to the data provided by the team. The dataset is several hundredth of terabytes big, so I work with the MapReduce framework, that allows me to navigate on an Hadoop cluster. I use Hive (SQL for Big Data) and R most of the time, but also Java, Python, Matlab and sometimes Stata.
    My results are regularly presented to the top management.

    - Results (for now):

    # Amadeus is now able to predict the popularity of any given travel, on any given platform.
    # The heuristic for maximizing the accuracy of the prices displayed on the end-user screen has been optimized thanks to greedy convex optimization algorithms under linear constraints and soft thresholding. This method is being patented.
    # I also gave trainings in Data Analytics (Machine Learning, supervised and unsupervised statistics) and R during the "Boston Academy".

    http://fr.slideshare.net/renuccif/data-analytics-session-1-2013-27929321





    Participation with 9 other teammates from Amadeus to the Big Data Hackathon in Boston, organized by Hack/Reduce : it is a coding competition, by team, that lasts 24 hours.

    Our project consisted in crawling the imdb database, spotting the locations and the release date of the most famous TV shows, linking each of those locations to the nearest and most relevant airport thanks to the geonames api, and studying the overall demand in flight tickets for this destination. By building a predictive model until the release date - 1 (univariate time series autoregression with lag estimation thanks to a cross-correlation criterion with hard thresholding), keeping the most relevant processes, and forecasting the future demand based on the model, we can evaluate the impact of the TV show itself.

    The results are amazing : we can now identify the TV shows that will increase the demand and quantify it. This has a huge business value for both the airlines and the online travel agencies : they can adapt prices (yield management) and advertise (event-driven marketing) to convert those extra searches into sales. We won the award for the "most creative, awesome use of data".

    See http://www.amadeus.com/nablog/2013/10/diary-of-a-hack-big-travel-data-hackathon/life-amadeus if you want to live the Hackathon with us !

  • Hemis Consulting - Co-founder, Consultant

    2012 - maintenant Co-founder of Hemis Consulting, consultancy support to the growth of start-ups and SMEs, on the following axis:

    - Business: Business Plan, to support the creation and fundraising,

    - Financial: audit and valuation, financing plans,

    - Technology: Business Intelligence, Data-Mining, technical feasibility study and design of structures.

  • INSEAD - Research in applied mathematics

    Fontainebleau 2012 - 2013 Research project in statistics and Machine Learning about "Leveraging Community Engagement for Brand Engagement", in conjunction with INSEAD researchers from all over the world, headed by Theodoros Evgeniou (head of the Decision Lab at INSEAD), leading to strategic recommendations for one of the biggest french internet portal.

    - Objectives :

    Maximizing the revenue of the advertising on one of the biggest french portal on the internet.

    - Framework :

    We worked with some of the best researchers in that field from the Insead Decision Sciences and Technology Management lab.

    - Results :

    We have built a model and delivered operational recommendations to the portal to increase its ad revenues, by improving the transformation rate between User Generated Content (low CTR, high number of visits) and Professional Generated Content (high and targeted CTR, low number of visits) of the website.

    - Learning :

    Working with and learning from academical researchers from all over the world (China, India and Greece).
    Building a link between theoretical results in applied mathematics and real-world question : some tradeoffs have to be made, between accuracy and efficiency.

    http://fr.slideshare.net/renuccif/leveraging-community-engagement-for-brand-engagement-2012-presentation
    http://fr.slideshare.net/renuccif/seminaire-2

  • Taupe Niveau - Entrepreneurship : iPhone applications "Taupe Niveau"

    2011 - 2011 - Objective :

    Development of an iPhone application suite "Taupe Niveau" (http://taupeniveau.com/) targeting the students in scientific CPGEs, gathering course material in mathematics, physics and chemistry, in a completely different way as usual : the concepts are presented here in a intuitive, creative and goal-oriented way, instead of the usual very formal and descriptive way.

    - Results :

    The apps are available on the AppStore :
    http://itunes.apple.com/us/app/mpsi-taupe-niveau/id487957345?mt=8
    (among others)

  • HeadLink Partners - Associate Strategy Consultant

    2011 - 2011 - Objectives :

    Change Management : Reengineering of a Corporate and Investment Banking worldwide, front to back, on several business lines.

    Development of the method of Change Management for Bank Efficiency Management practice, training sessions for consultants, development of commercial offer for operational excellence in retail banking and investment banking.

    - Framework :

    Team of 4 Consultants (including a senior manager), and several hundredth of employees impacted by the mission.

    - Results :

    The method is implemented (tools used during the mission), functional training, and business proposal updated. The tools have been field-tested during the mission to which I was attached, on the equity line.

    - Learnings :

    Teamwork with very experienced people, dealing with different skills / experience, clarifying reasonings, structuring a method of problem solving, understanding business problematics of a big organization.

  • Orange Healthcare - Entrepreneurship : Introduction of hotel services in hospitals

    Paris 2010 - 2011 - Objective :

    Introduction of hotel services in hospitals.

    - Framework :

    Collaboration with the R&D branch of Orange Healthcare. Team of 10 + 1 tutor, from three different schools (Central Paris, ESSEC Business School, Strate College Designers), together with the CPI program (Creation of an Innovative Product).
    Multicultural team, whose paradigms oppose and complement each other:
    # Engineer: "What is possible?"
    # Commercial: "Where is the value?"
    # Designer: "How emotion rises?"
    At the intersection is the result of our project.

    - Results :

    Project presented to the Executive Committee of Orange Healthcare. It will be tested in hospitals in a few years (2015).


    - Learnings :

    Combining different approaches to three visions, realizing a project by managing all aspects of the outcome technical, financial and aesthetic).

  • Medialis - Entrepreneurship : Creating a website selling online gerontechnology products

    Corbas 2009 - 2010 - Objective :

    Creating a website selling gerontechnology products : technologies that improve the living conditions of our seniors.

    - Framework :

    Team of five students, including two foreigners, during 9 months.

    - Results :

    The website is implemented, functional, and will be online as soon as enough contracts with manufacturers will be signed (in order to start with a fairly comprehensive catalog).


    - Learnings :

    Learning team dynamic, roles allocation and coordination. The approach of different cultures has allowed me to develop my interpersonal skills.

Formations

  • Ecole Normale Supérieure

    Paris 2012 - 2013 Master's degree in research in applied mathematics (Machine Learning, artificial intelligence), Applied mathematics, Machine Learning

    The master's degree "MVA", based in the Department of Mathematics at the Ecole Normale Supérieure, forms dozens of students from the french Grandes Ecoles to careers in public and private research in the field of applied mathematics, among other things, statistical learning.

  • ESSEC Business School

    Cergy 2011 - 2012 Master of Business Administration (MBA)

  • Ecole Centrale De Paris

    Chatenay Malabry 2009 - maintenant Master's degree in engineering from the Ecole Centrale Paris

    Major : applied mathematics/Machine Learning, track : Strategy/Marketing/Finance.

  • Lycée Massena

    Nice 2006 - 2009

  • Lycée Guillaume Apollinaire

    Nice 2003 - 2006 BAC S option Maths

    , mention TB

Réseau

Annuaire des membres :