Looking to transition into a role as a data professional? This talk aims to -
1) Demystify data professional roles in the industry
2) Learn strategies to land your first job as a data professional
3) Make you fall in love with the process of learning!
2. Kickstarting your
career in AI
Priyanka Kulkarni
Senior Machine Learning Scientist
Office Experience Organization
linkedin.com/in/priyankakulkarni11/
3. 3
Session goals
Demystify data professional roles in the industry
Learn strategies to land your first job as a data professional
Make you fall in love with the process of learning!
4. 4
What role do you want to play?
Career Core Activity Core Skills
Machine Learning Researcher Advance human knowledge
Scientific method, math, basic
programming
Data Scientist Stories from data
Statistics, data manipulation,
communication
(Applied) Machine Learning Scientist Build predictive models
Machine learning algorithms, domain
specific feature engineering, basic
programming
Machine Learning Engineer
Integrate machine learning into
systems
Software engineering, conceptual
machine learning
Machine Learning Architecture /
Program Manager
Design solutions that leverage
machine learning
Software design skills, customer
empathy, Strong conceptual machine
learning
Credits: Geoff Hulten Blog | Subscribe
5. 5
Evaluate : Now vs role.next()
Identify the
delta
Evaluate
Which roles
energize you?
aka.ms/careers
is your friend!
Domain : Ex. Healthcare, advertising, security
Basic requirements : Ex. Coding in python/R,
knowledge of statistics and basic ML algorithms
Specialization : Ex. Experience with distributed
processing, privacy preserving ML
6. Essential skills
Analysis
• Extraction
• Wrangling
• Statistical
reasoning
• Understand
the business
Imagination &
Data manipulation
• Pick a
language
• Ask questions
• Use data to
answer them
• Data thinking
vs data
reporting
Visualization
• Translate data
into
consumable
visuals
• Take that
visualization
course /
tutorial / book
Communication
• Explain your
results
• Numbers
Implications
Decisions
• Use data to
build trust
7. 7
Learningframework
• Alice and Bob want to dedicate 2 hours per week with the goal to
build their first deep learning model for image classification
Decide on a learning budget
and outcome
• Bob learns best with structure and strict timelines, so they prefer a
part time masters vs Alice enjoys the flexibility of learning from blogsFormal vs Informal learning
• Bob identifies the following specific topics - working with image data,
evaluation techniques, CNNs
Backtrack - identify specific
topics
• Alice wants to implement their models from scratch and wished to
fully comprehend the underling mathDecide - Deep vs Wide
• Alice aggregates all the relevant blogs and books they plan to read
and starts with the fundamentalsGather resources
• Bob likes to take one class per week and dedicate 2 hours for
assignments
Fit them as per your learning
budget
• Because Alice & Bob already have!
Start now!
8. Art of assimilating research papers
Credits: Andrew Ng
Assimilating research
papers
Read
title/abstract/figures
(first pass)
Intro + Conclusion +
figures + skim the
rest (skip related
works)
Read the paper but
skip the math.
Read the whole
thing but skip parts
that don’t make
sense
1.What did the
authors try to
accomplish and what
were the key
elements of the
approach
What can you use
yourself ?
9. Art of assimilating research papers
Assimilating research
papers
Read
title/abstract/figures
(first pass)
Intro + Conclusion +
figures + skim the
rest (skip related
works)
Read the paper but
skip the math.
Read the whole
thing but skip parts
that don’t make
sense
1.What did the
authors try to
accomplish and what
were the key
elements of the
approach
What can you use
yourself ?
10. Art of assimilating research papers
Assimilating research
papers
Read
title/abstract/figures
(first pass)
Intro + Conclusion +
figures + skim the
rest (skip related
works)
Read the paper but
skip the math.
Read the whole
thing but skip parts
that don’t make
sense
1.What did the
authors try to
accomplish and what
were the key
elements of the
approach
What can you use
yourself ?
11. Art of assimilating research papers
Assimilating research
papers
Read
title/abstract/figures
(first pass)
Intro + Conclusion +
figures + skim the
rest (skip related
works)
Read the paper but
skip the math.
Read the whole
thing but skip parts
that don’t make
sense
1.What did the
authors try to
accomplish and what
were the key
elements of the
approach
What can you use
yourself ?
12. Art of assimilating research papers
Assimilating research
papers
Read
title/abstract/figures
(first pass)
Intro + Conclusion +
figures + skim the
rest (skip related
works)
Read the paper but
skip the math.
Read the whole
thing but skip parts
that don’t make
sense
1.What did the
authors try to
accomplish and what
were the key
elements of the
approach
What can you use
yourself ?
13. Keep learning by building projects
Apply skills
immediately
• Ex: just read about
pandas indexing ?
Take a dummy
dataset, try it out
now
• Following a
statistics textbook?
Try to solve exercise
problems with code
Join a competition /
solve real word
problem
• Download that
dataset and
attempt modeling
• Try different
evaluation methods
and understand
when to use them
• Go back to the
math later (but
don’t skip it!)
Learn Software
Engineering best
practices
• Gradually learn how
to build scripts for
your data science
workflow
• Keep track of your
code using source
control tools
• Put effort in
structuring your
code
14. 14
Be a part of the AI scene
Subscribe to receive regular
AI updates – it’s free
(mostly) and invaluable!
Identify ML / Data Science
groups – inside and outside
Microsoft
Attend at least one
conference per year – and
share your learnings
Find a mentor!
15. 15
Opportunitiesfortransition
Create
ML opportunities within
your current role/ project
•Start your own ML side
project & recruit
Make the most of
hackathons
Identify
Data Science roles in your
current domain/
organization
Roles that excite you
across the industry
Evaluate
Domain + requirements
+ specialization
Informational session
with the hiring manager
Talk to folks who already
have the role you aspire
towards
16. 16
Key learnings and insights
Identify the
role you’d
ideally want
to play
Evaluate –
now vs next
role
Upskill
Identify &
create
opportunities
Get that role!
This slide is required. Do NOT delete. This should be the first slide after your Title Slide.
This slide should describe what your goals are for this session. This information lets your audience know what you are trying to accomplish with your talk or tutorial—i.e., what value will attendees get by investing their time listening to you.
You should not spend more than 1 minute presenting this slide.
General examples of session goals could be (you will have to create your own specific goals):
Introduce a new technique or approach to solve a customer problem
Compare two approaches and explain why one is superior
Describe a project and the learnings that audience members can apply from it
Teach audience members how to use a specific technology