Life in academia and elsewhere Sophie Cluet Directorate General for Research and Innovation Mathematics; Physic; Nanoscience and Nanotechnologies; Security; Information and Communication Science and Technology
Dec 21, 2015
Life in academia and elsewhere
Sophie CluetDirectorate General for Research and Innovation
Mathematics; Physic; Nanoscience and Nanotechnologies; Security; Information and Communication Science and
Technology
Organization
Life as a CS researcher Life as a CTO Life as a research center manager Life as a science policy « maker » in a
Ministry
PhD
Research, including • Reading and writing papers • Going to conferences, workshops, …• Eventually, system development and experimentation
Teaching Collective interest tasks
• Team (annual reports, seminars, web site, …)• Institution (communication, various committees, …)• Community (reviewing, …)
Advising Master students
GrantsEventually, consulting for industry
And the normal life: family, friends, hobbies, sports…
Junior
Research, including • Reading and writing papers • Going to conferences, workshops, …• Eventually, system development and experimentation
Teaching Collective interest tasks
• Team (annual reports, seminars, web site, …)• Institution (communication, various committees, …)• Community (reviewing, …)
Advising PhD or Master students Grants Eventually, consulting for industry
And the normal life: family, friends, hobbies, sports…
Senior
Research, including • Reading and writing papers • Going to conferences, workshops, …• Eventually, system development and experimentation
Teaching Collective interest tasks
• Team (annual reports, seminars, web site, …)• Organization (communication, various committees, …)• Community (reviewing, …)
Advising (PhD students, etc.) Grants Eventually, consulting for industry
And the normal life: family, friends, hobbies, sports…
Life as a CS ResearcherE
As for men, an exciting life :
Relative freedom
General feeling of getting smarter
Mainly smart colleagues
Great highs and lows
Part of an international community
…
Life as a CS ResearcherE (cont.)
Differences :
Pregnancies : one of the jobs where it is not really a handicap
Still, part of a minority• Negative effects : getting better in France but still a tendancy
to consider women less seriously than men (cf. Françoise Giroud "La femme sera vraiment l'égale de l'homme le jour où, à un poste important, on désignera une femme incompétente.") => women who succeed usually work harder than their male colleagues
• Positive effects : women are more noticable => Easier to make friends, more offers of responsabilities (program committees, …
Why did I quit ?
Essentially, the opportunity of a new challenge
But also, after 12 years, feelings of YAP and YAC
Xyleme History
1999 : a bet 1999 – 2000 : getting ready 2000 : starting the company 2001 : the first customers come in 2002 : getting more money and what
comes with it – end of the story 2007 : end of Xyleme
1999 : The Bet
An audit of the team’s work by an old friend (François Bancilhon) Semistructured data (among which XML) :
models, query languages, query processing Web applications : crawlers, integration
(wrappers/mediators) with light language analysis, distributed query processing, …
The old friend’s bet : a year to build the new Alexandrian library Hypothesis : soon, the interesting data of the
Web will be in « good » XML (i.e., not XHTML)
1999-2000 : getting ready
Working hard on a product and a business model with the help of INRIA Alliance with an AI team Hiring a technical director and some engineers 2 seminars with friends from industry and capital
risk Deciding who will be who : founders, boss,
directors (business development, finance & administration, commercial, technical, …)
Getting money (before the internet bubble) from friendly capital riskers
2000 : starting the company
More Hiring, altogether around 20 people mostly engineers
Getting a place to work and what goes with it (from computers to cleaning staff)
Madly crawling the web to find near to 0 interesting XML data …
Looking for a CEO (got one in 2001)
2001 : The First Customers
Change of the business model From service provider to software editor From public to private data
The main targets : the press and the banks who have tons of XML or XMLizable data
Small contracts, few money and a lot of work on pilots
2002 : Getting More Moneyand what comes with it
A great technology, INRIA in the background, a few satisfied customers, a convincing team => lots of money but this time not only from friends
More customers but hard to get : more and more pilots and specialized development
Why did I quit ?And some tips…
Another exciting opportunity, but mainly : having a CTO and a technical manager in a startup is a bad idea…
Some other mistakes : Leaving the lab before having validated the idea Spending a lot before having the customers A technology rather than something a decision maker can
understand A weak management
Still, a great experience : Real life Many different jobs in one Happy customers …
INRIA Rocquencourt
The largest of the 6 (now 8) INRIA’s Research CentersBudget (excluding civil servants salaries) : 11 M€Around 500 people• 350 scientists in 40 research project teams (CS and applied mathematics)
• 130 « long » time researchers• 20 postdocs• 120 PhD• 50 Master • 50 engineers
• 150 TA in 8 support services • Administration and finance• Communication• Human resources• Industrial relationship and technology transfer • Project assistants• Restaurant • Site logistic• Technical support
Job Description
By the CEO :1. Scientific leader2. Member of the INRIA Board of directors3. CEO of a SME
In reality : reverse order
How did I spend my time ? 106 days/year of recurrent meetings from 2 to many people (46, 40,
20) 50 days/year of e-mails (under evaluation - basis 5mn per sent
message) ~ 300 days/year
Other meetings (to solve problems – often HR, to work on strategy, to convince people, to treat various demands, …)
Preparing the meetings Writing or re-writing communications
It’s all about people !
That’s why it’s great That’s why it’s time consuming That’s why you cannot reach perfection
“Happiness” and a feeling of
belonging are essential
It’s all about people !
That’s why it’s great That’s why it’s time consuming That’s why you cannot reach perfection
“Happiness” and a feeling of
belonging are essential
• Project team• Service• Research center• General scientific direction• General “functional” direction• Institution
It’s all about people !
That’s why it’s great That’s why it’s time consuming That’s why you cannot reach perfection
“Happiness” and a feeling of
belonging are essential
• Project team• Service• Research center• General scientific direction• General “functional” direction• Institution
Not so far from research as one could think …
Understand the big picture Identify/analyze the challenges For each selected one :
Search for and define the right model Implement Evaluate Iterative small or large delta improvements
Some of the challengesscientific leader
The basic one : every year, sharing budget and other resources among project teams => scientific priorities help
Defining scientific priorities (every 4 years or so – although, obviously, no big changes from one exercise to the next) Why define priorities? What is the best process? Why is a priority a priority? Is there a life outside a priority?
Some of the challengesMember of the board of directors
CEO decisions impact the institution various structures differently. 2 examples : Scientific advisor status Information system
Every decision has an impact, but not all of them interest you… They should. Also, you may not be aware of all of them… You should.
What is needed ? Keeping a bottom-up top-down bottom-up top-down
bottom-up top-down … persistent flow of information An heavy touch of transversal communication A “happy” nature and a big mouth
Some of the challenges CEO of a SME
The main one : defining a strategy for the whole (projects and services) within that of the institution
The restaurant People love it Requires about 15 employees Less than half are civil servants, half of which will retire soon No possibility to recruit on more than a 9 months basis (or recruit
less of some other categories) hard on the civil servants
More than 100 newcomers each year Scientific, administrative, technical From 6 months to 4 decades For some, weak recruiting process (public rules) All managers are not “managers”
Why did I quit And how did I find my next job
Why did I quit Instability among INRIA’s CEOs and/thus(?) I
was tired Still, maybe my favorite job so far !
How did I find my next job Networking In 2006, in France, in this area, being a woman
helps
Life as « Policy Maker » in a Ministry
2006 – 2008French Ministry of Higher Education and Research
Directorate General for Research and InnovationA3 department : Mathematics; Physic; Nanoscience and Nanotechnologies;
Security; Information and Communication Science and Technology
Service de l'innovationet de l'action régionale
Sciences de la Terre et de l'univers,géo-environnement, aéronautique,
transport, Espace
DIRECTION GÉNÉRALE DE LA RECHERCHE ET DE L'INNOVATION
Bureau de la recherche etdéveloppement en
entreprise
Bureau valorisation,propriété intellectuelle
et partenariat
Bureau de la création et du développement
des entreprises technologiques
Bureau de l'action régionale
Bureau de la réglementation
et des statuts
Bureau politique contractuelle
et coordination de la tutelle
Bureau de programmation des
moyens et des TGI
Bureau des affaires européennes
Mission emploi scientifique
Mission de l'information et de la culture
scientifiques et techniques
Mission Parité
Chimie, SPI, physique nucléaireet des hautes énergies, énergie,
développement durable
Mathématiques, physique, nanos,usages, sécurité, STIC
Biotechnologies,ressources, agronomie
Santé
Sciences de l'hommeet de la société
Sous-direction de l'appui à la tutelle
et affaires européennes
Département des étudeset de la prospective
Département des politiquesde recherche et d'innovation
Mission coordinationde la MIRES
et ystèmes d'information
DRRT
Direction de la stratégie
Organigramme de la DGRI
The « canonical » National Research and Innovation System
Three levels1. Orientation, national policy (macro-objectives)2. Programmation (translation of macro-objectives in programs)3. Operation (research operators)
The principles Autonomy of the three levels (distinct institutions) Interactions defined by 4 to 5-years contracts Each level is evaluated (ex post) by the one above
Job description
Service to the Minister Definition of the national R&I strategy (A3 domains) Implementation and follow-up of the strategy :
Supervision of level 2 • ANR (« matière et information »)• CEA (DRT, DSM)• CNRS (MPPU, ST2I)• GENCI• INRIA• Others (foundations, public interest groups, …)
Scientific policy of research operators (level 3 !) Collaboration with other ministries
European programs (ICT, PERS, NMP) (Too) many assessments/evaluations (historical, should slow
down)
National vs. Research Center strategy
The reasons for defining a strategy are more or less the same, but: The stakes are much larger (20 B€ spent by the “MIRES” each year,
desired impact on the economy) The “domain” is larger, more interconnections The grain is larger (you don’t see all the individuals) The decision circuit is much longer and tougher
The method is also more or less the same, some try to formalize it (science of science policy)
The method I’d like to say I use (currently)
(Bottom-up/top-down cycles + a touch of transversality + a lot of reading) to identify/order the domain main challenges (continuous process)
For each challenge1. (SWOT) analysis (can be done by consultants – at this level,
indicators are a must)• Validate the opportunity to act• Find clues on what to do
2. Collective work (all levels) to devise some action scenario(s) (and prepare the implementation)
3. Eventually, analysis to choose one of the scenarios4. Sell the scenario to the politics5. Implement
You think research in CS is a long process? Try strategy
Some A3 challenges
“The French mathematics school: danger ahead because of retirements, lack of students, new policy for HE&R employment combined with universities new autonomy
Improvement of France’s performances in some identified sectors: Computer security Ambient intelligence Nanotechnologies for environment or health …
National structuring of some sectors: Electronics Grids (production and research)
Improvement of the management culture in the HE&R institutions : an emergency given the many changes that have occurred or are occurring
ICST employment : danger ahead for research because of many students and many R&D projects (workload can be twice that of other disciplines)
…