last modified: 2017-10-02
The goal for this group project is to make you go and meet experts in one of the many topics discussed in this course.
Preparing the interview, you will get to learn the topic you choose more in depth.
Meeting the expert, discussing with her, and learning about her environment (company, colleagues, …), you will understand better what "doing big data" means in practice.
Recording and publishing the interview will create a public outcome that you can add to your portfolio for your job search.
Groups can have 1 to 3 members.
pick a topic
find an expert to interview
prepare the interview
record the interview
post the interview online
Send me an email or meet me during office hours to validate the topic you choose.
Write the topic you choose in the spreadsheet where you formed groups (see above)
Examples of tech-y topics:
Why are GPUs (graphic cards) used in machine learning and how do you use them in your job?
Practising text mining in … (pick a company here): methods and outcomes
Managing Petabytes of data at … (pick an expert at a company like LeBonCoin, Airbus, etc.)
Examples of business-y topics:
How does data science transform the job of the consultant in … [pick a practice]?
How does robotics change supply chains
AI for the care of the elderly: hype or near future?
Through your personal networks, the emlyon alumni network, … you can find an expert able to give you a bit of time to answer questions.
Often, you have an expert in mind before defining a topic. This is fine, you can elaborate and refine a topic based on the expert you find.
The expert does not need to be based in France, as long as you can interview this person with a good audio. If that’s the case, triple check that this person has access to a good microphone: long distance interviews with an horrible audio don’t make a podcast!
Preparing the interview is essential. A good format is a semi-structured interview: you should have a list of questions to ask, but be able and ready to open new directions if they arise during the interview.
To achieve this, before the interview you must:
learn about the subject you will focus on
learn about the interviewee and the industry / organization she’s embedded in
You can share your questions with the expert before the interview, to make sure they see what you expect from them in terms of knowledge domain and degree of precision.
The interview itself has a couple of keypoints:
Make sure you have a good microphone. TEST EVERYTHING BEFORE. DOUBLE CHECK YOUR BATTERIES. If you want to do an audio + video recording, it is definitely possible as well.
conducting the interview:
Try to target an interview lasting 20 to 30 minutes. Start with questions where the expert introduces herself: name, position, organisation at least. If you feel the interviewee stays too long on an answer, try orienting the conversation to the next question - as fluidly as you can!
See the syllabus for the deadline to publish the podcast. Grading is as follows:
33% of the grade on the quality of the questions (showing the interviewer has a good knowledge of the topic, and helping the expert discuss interesting topics)
33% of the grade on the conduct of the interview (good intro, good management of the time, no blanks…)
33% of the grade on the quality of the podcast (sound of the interviewer and interviewee)
Can the podcast be in French?
Yes, the podcast can be in French or English.
Can I use my smartphone to record the interview?
Yes, as long as it records good audio. Test it. You could do video as well with your smartphone. Just check the quality.
I have found this person… is she an expert?
An expert can be young or not, from a big company or not, famous or not. This is a person who, through its current or past activities, can demonstrate a very good knowledge of the topic she is discussing.
I am not sure this topic is what you expect
Choose a topic related to one of the topics discussed in this course. Send me an email or meet me during office hours to validate it.
Can I interview several experts at the same time?
Yes. You might meet a pair of co-founders, or a team of data scientists… it is up to you to keep an orderly interview, which can be harder (multiple voices…). But it can be definitely interesting!
We are two or three in our group. Should we all talk and ask questions in the podcast?
All members of the group should participate in preparing the podcast. One person can conduct the interview. If you listen to this podcast, you will see that interviews by 2 persons can work very well.
Do you have another question? Send it at email@example.com
Find references for this lesson, and other lessons, here.
This course is made by Clement Levallois.
Discover my other courses in data / tech for business: http://www.clementlevallois.net
Or get in touch via Twitter: @seinecle