CRAFTING A COMPELLING DATA
SCIENCE RESUME
Jennifer Hay
IT Resume Writer
IT Resume Service
Arushi Prakash
Data Scientist II
Zulily
Women in Data Science Conference
Puget Sound
May 4th, 2020
JENNIFER HAY
4
IT Resume writer, 2009 – Present
Specializing in writing resumes and
Linked Profiles for data analysts, data
scientists, data engineers, data
stewards, etc.
Writes data and information
management exams for
eLearningCurve.
ARUSHI PRAKASH
Data Scientist @ Zulily, 2019 - Present
Works on recommender systems,
natural language processing, and A/B
testing
Education
PhD, Chemical Engineering, 2018
BS, Chemical Engineering, 2014
5
WE WANT TO HELP YOU
1. Best Practices
2. Job Profiles in Data
Science
3. What is the purpose
of storytelling in a
resume?
6
TODAY’S AGENDA
Resume
Interviews
Job
Job portals
LinkedIn
Recruiters
Staffing agents
Recruiters
Hiring Managers
Future Co-workers
BEST PRACTICES
FOR RESUME WRITING
Including nuggets like “Always keep to a one-
page resume!” and more …
7
© 2020 Arushi Prakash and Jennifer Hay 8
Format
• Keep it plain and human readable
• Chronological vs. Functional
Writing
• Keep relevant work experiences
• Use keywords from job posts
Writing style
• Use spelling and grammar checkers
• Avoid generic statement
UNDERSTAND THE JOB PROFILE
9
THE FIELD OF DATA
SCIENCE
Diverse
Nascent
Lacks definition
10
Computer
Science
Data
Mathematics &
Statistics
Machine
Learning
Engineer
Data
Engineer
Data Analyst
Data Scientist
DATA ANALYST ROLES
11
DATA SCIENTIST ROLES
12
MACHINE LEARNING ENGINEER ROLES
13
DATA ENGINEER ROLES
14
WHAT DEGREES DO THEY HAVE?
https://medium.com/indeed-engineering/where-do-data-scientists-come-from-
fc526023ace
15
WHAT DEGREES DO THEY HAVE?
https://medium.com/indeed-engineering/where-do-data-scientists-come-from-
fc526023ace
16
GENDER BIAS IN JOB
DESCRIPTIONS
17
APPLYING FOR JOBS
Women only apply for jobs if
they meet 100% of the
requirements.
Men apply if they meet 60%.
18
https://hbr.org/2014/08/why-women-dont-apply-for-jobs-
unless-theyre-100-qualified
THIS ISSUE IS WIDESPREAD
19
https://www.ziprecruiter.com/blog/removing-
gendered-keywords-gets-you-more-applicants/
THE IMPORTANCE OF
STORYTELLING
20
Carol’s background
• Conducting economic research and
publishing articles.
• Attending and presenting research findings at
leading conferences.
• Reviewing and critiquing articles written by
other researchers.
• Teaching undergraduate and graduate
economic courses.
Many of my clients focus on the
technology that they delivered. While
this is important, it is also essential to
think about the journey to getting to
those solutions.
I find that it is often the journey that
provides the most compelling
information and brings out a person’s
leadership, management, and technical
expertise.
21
FRAMEWORK FOR
STORYTELLING
FRAMEWORK FOR
STORYTELLING
• Based on my analytics experience, invited to
speak at WiDS, an organization that inspires
and educates data scientists worldwide.
o Engaged attendees in learning about the
use of data science in economics.
o Received very positive feedback.
o Presented one of my research projects
and performed exercises to illustrate
regression analysis, computational
modelling, and machine learning.
22
FRAMEWORK FOR
STORYTELLING
FRAMEWORK FOR
STORYTELLING
Reinforce information.
Have you done any of the following:
• Mentored someone through a problem?
• Trained colleagues to use new technologies
or technical practices?
• Given a presentation about your work?
• Advocated for needed improvements?
• These are just few examples.
23
FRAMEWORK FOR
STORYTELLING
FRAMEWORK FOR
STORYTELLING
FRAMEWORK FOR
STORYTELLING
24
Who got
value and
how?
What was
produced?
What were
the
challenges?
What did
you do?
How did
you do it?
What did
you
learn?
value of story elements
low high
WHAT DID YOU DO?
• Analytics Lifecycle: data discovery, data
aggregation, planning of the data model,
data model execution, communication of
the results, and operationalization.
• Project Management Lifecycle:
project initiation, planning, execution, and
closure.
• Software Development Lifecycle:
planning, designing, implementing/coding,
testing, deploying, and maintaining.
• Product Development Lifecycle:
concept, user research, feasibility and
proof of concept, architecture and design,
implementation, test and release,
manufacturing, and support.
Step back from your specific work and
think about the overall area in which you
were working. Who is part of your team?
25
WHAT DID YOU DO?
Created a research process and defined each
step for collaborative teamwork. Taught my
team to:
(1) Ask a meaningful question, (2) Find suitable
data sets, (3) Decide on an appropriate
statistical method, (4) Use data analytics
software to conduct analyses, (5) Answer the
question based on quantitative results, and (6)
Communicate findings both in a written report
and an oral presentation.
Step back from your specific work and
think about the overall area in which you
were working. Who is part of your team?
26
HOW DID YOU DO IT?
Modernized a bioinformatics environment to
strengthen their primary product and enable
customers to conduct advanced research.
• Converted to a Django framework to enable
rapid development with a large resource
base and to provide a scalable web
framework to meet our growth
expectations.
• Further enhanced the Django application
with AngularJS on the client-side and added a
variety of new features (e.g.“shopping cart
functionality”) to a Drupal based information
system.
Think about the frameworks, tools, methods, and
algorithms you used to achieve the outcome.
27
WHAT WERE THE
CHALLENGES?
• Short deadlines
• Resource constraints and a lack of in-
house talent
• Significant data quality problems and data
access problems
• A company new to data science with high
expectations for what DS could do
• Technical problems
• Leadership/organizational problems
• A poor data culture and a lack of data
literacy within the company
• A requirement that the data scientist also
be the data engineer
• Data compliance and privacy issues
Think about what made the task difficult to
achieve.
28
WHAT WERE THE
CHALLENGES?
• Resolved a critical problem in a cloud-based
Human Capital Management (HCM) product
with a soft release scheduled at a major
conference.
• Results: The company filed a patent and my
innovative solution will be used for other
word embedding models.
• Challenges: Model updates required
retraining the whole dataset. Also, it would
violate our customer data policy by merging
customer data and the company’s hiring data.
• Created and tested a new online training
algorithm, which reduced 95% of model
updating time and prevented potential data
security issues in production.
Think about what made the task difficult to
achieve.
29
WHAT DID YOU LEARN?
• Learned the importance of data quality in
analytic processes during an AI Claim
Lifecycle project to predict the likelihood for
a successful audit of an insurance claim.
o Challenge: Non-standardized claims
created a data integration problem.
o Selected a binary classifier logistic
regression model as the best predictor.
o Defined a new workflow for metadata and
implemented a process upstream to
standardize fields.
• Gained experience in the software side and
the ML side of building production scale AI
applications, using distributed Apache Spark
frameworks, to score very high data
volumes.
This is particularly important for those new
to data science.
30
WHAT WAS PRODUCED?
• Developed a tool that tracks customer
behaviors and purchasing at a national
grocery chain.
o Designed and created a data pipeline that
included geospatial data for a broader
understanding of the entire region.
o Enabled each business division to analyze
and quickly respond to customer needs
due to the evolving changes in supply
chains.What did you create?Was it a single tool, a
portfolio of services, or an analytic platform?
31
If you work in industries that have
been significantly impacted by the
Coronavirus, now is the time to think
about what analytic products can meet
these evolving needs.
WHO GOT VALUE AND
HOW?
• Advise undergraduate and graduate
students, mentoring through the process
of determining a career path by channeling
their commitment and passion for
research and analysis.
o Leveraged a student's interest in the
local business environment by
coauthoring a paper on "Startups in Los
Angeles.”
o Resulted in the student getting a
Research Assistant job at the L.A.
County Offices.
o Received the 2019 Outstanding
Graduate Student Research and
Mentorship Award at my college based
on her nomination.
Think about who got value from the success of an
effort or project. Demonstrate and quantify
results here.
32
RESUME SUMMARIES
A Data Engineer and a Database Engineer with
extensive project management experience as a
Certified PMP, creates major data-driven value
opportunities for the company.
• Turns around troubled projects and
reenergizes the team to regain commitment.
• As an experienced financial analyst, uses data
storytelling techniques to create informative
graphs with comprehensive narratives.
• Works collaboratively throughout the
analytics lifecycle and has particular
expertise in communicating and
operationalizing the results.
In resume summaries it is also important to
think about the issues that we’ve covered.
33
SUMMING IT ALL UP
34
LESSONS IN RESUME WRITING
• Follow best practices
• Know which role you are
aiming for
• Tell a coherent story with
what, why, and how
components
• Reach out to people
35

Crafting a Compelling Data Science Resume

  • 1.
    CRAFTING A COMPELLINGDATA SCIENCE RESUME Jennifer Hay IT Resume Writer IT Resume Service Arushi Prakash Data Scientist II Zulily Women in Data Science Conference Puget Sound May 4th, 2020
  • 2.
    JENNIFER HAY 4 IT Resumewriter, 2009 – Present Specializing in writing resumes and Linked Profiles for data analysts, data scientists, data engineers, data stewards, etc. Writes data and information management exams for eLearningCurve.
  • 3.
    ARUSHI PRAKASH Data Scientist@ Zulily, 2019 - Present Works on recommender systems, natural language processing, and A/B testing Education PhD, Chemical Engineering, 2018 BS, Chemical Engineering, 2014 5
  • 4.
    WE WANT TOHELP YOU 1. Best Practices 2. Job Profiles in Data Science 3. What is the purpose of storytelling in a resume? 6 TODAY’S AGENDA Resume Interviews Job Job portals LinkedIn Recruiters Staffing agents Recruiters Hiring Managers Future Co-workers
  • 5.
    BEST PRACTICES FOR RESUMEWRITING Including nuggets like “Always keep to a one- page resume!” and more … 7
  • 6.
    © 2020 ArushiPrakash and Jennifer Hay 8 Format • Keep it plain and human readable • Chronological vs. Functional Writing • Keep relevant work experiences • Use keywords from job posts Writing style • Use spelling and grammar checkers • Avoid generic statement
  • 7.
  • 8.
    THE FIELD OFDATA SCIENCE Diverse Nascent Lacks definition 10 Computer Science Data Mathematics & Statistics Machine Learning Engineer Data Engineer Data Analyst Data Scientist
  • 9.
  • 10.
  • 11.
  • 12.
  • 13.
    WHAT DEGREES DOTHEY HAVE? https://medium.com/indeed-engineering/where-do-data-scientists-come-from- fc526023ace 15
  • 14.
    WHAT DEGREES DOTHEY HAVE? https://medium.com/indeed-engineering/where-do-data-scientists-come-from- fc526023ace 16
  • 15.
    GENDER BIAS INJOB DESCRIPTIONS 17
  • 16.
    APPLYING FOR JOBS Womenonly apply for jobs if they meet 100% of the requirements. Men apply if they meet 60%. 18 https://hbr.org/2014/08/why-women-dont-apply-for-jobs- unless-theyre-100-qualified
  • 17.
    THIS ISSUE ISWIDESPREAD 19 https://www.ziprecruiter.com/blog/removing- gendered-keywords-gets-you-more-applicants/
  • 18.
  • 19.
    Carol’s background • Conductingeconomic research and publishing articles. • Attending and presenting research findings at leading conferences. • Reviewing and critiquing articles written by other researchers. • Teaching undergraduate and graduate economic courses. Many of my clients focus on the technology that they delivered. While this is important, it is also essential to think about the journey to getting to those solutions. I find that it is often the journey that provides the most compelling information and brings out a person’s leadership, management, and technical expertise. 21 FRAMEWORK FOR STORYTELLING FRAMEWORK FOR STORYTELLING
  • 20.
    • Based onmy analytics experience, invited to speak at WiDS, an organization that inspires and educates data scientists worldwide. o Engaged attendees in learning about the use of data science in economics. o Received very positive feedback. o Presented one of my research projects and performed exercises to illustrate regression analysis, computational modelling, and machine learning. 22 FRAMEWORK FOR STORYTELLING FRAMEWORK FOR STORYTELLING Reinforce information.
  • 21.
    Have you doneany of the following: • Mentored someone through a problem? • Trained colleagues to use new technologies or technical practices? • Given a presentation about your work? • Advocated for needed improvements? • These are just few examples. 23 FRAMEWORK FOR STORYTELLING FRAMEWORK FOR STORYTELLING
  • 22.
    FRAMEWORK FOR STORYTELLING 24 Who got valueand how? What was produced? What were the challenges? What did you do? How did you do it? What did you learn? value of story elements low high
  • 23.
    WHAT DID YOUDO? • Analytics Lifecycle: data discovery, data aggregation, planning of the data model, data model execution, communication of the results, and operationalization. • Project Management Lifecycle: project initiation, planning, execution, and closure. • Software Development Lifecycle: planning, designing, implementing/coding, testing, deploying, and maintaining. • Product Development Lifecycle: concept, user research, feasibility and proof of concept, architecture and design, implementation, test and release, manufacturing, and support. Step back from your specific work and think about the overall area in which you were working. Who is part of your team? 25
  • 24.
    WHAT DID YOUDO? Created a research process and defined each step for collaborative teamwork. Taught my team to: (1) Ask a meaningful question, (2) Find suitable data sets, (3) Decide on an appropriate statistical method, (4) Use data analytics software to conduct analyses, (5) Answer the question based on quantitative results, and (6) Communicate findings both in a written report and an oral presentation. Step back from your specific work and think about the overall area in which you were working. Who is part of your team? 26
  • 25.
    HOW DID YOUDO IT? Modernized a bioinformatics environment to strengthen their primary product and enable customers to conduct advanced research. • Converted to a Django framework to enable rapid development with a large resource base and to provide a scalable web framework to meet our growth expectations. • Further enhanced the Django application with AngularJS on the client-side and added a variety of new features (e.g.“shopping cart functionality”) to a Drupal based information system. Think about the frameworks, tools, methods, and algorithms you used to achieve the outcome. 27
  • 26.
    WHAT WERE THE CHALLENGES? •Short deadlines • Resource constraints and a lack of in- house talent • Significant data quality problems and data access problems • A company new to data science with high expectations for what DS could do • Technical problems • Leadership/organizational problems • A poor data culture and a lack of data literacy within the company • A requirement that the data scientist also be the data engineer • Data compliance and privacy issues Think about what made the task difficult to achieve. 28
  • 27.
    WHAT WERE THE CHALLENGES? •Resolved a critical problem in a cloud-based Human Capital Management (HCM) product with a soft release scheduled at a major conference. • Results: The company filed a patent and my innovative solution will be used for other word embedding models. • Challenges: Model updates required retraining the whole dataset. Also, it would violate our customer data policy by merging customer data and the company’s hiring data. • Created and tested a new online training algorithm, which reduced 95% of model updating time and prevented potential data security issues in production. Think about what made the task difficult to achieve. 29
  • 28.
    WHAT DID YOULEARN? • Learned the importance of data quality in analytic processes during an AI Claim Lifecycle project to predict the likelihood for a successful audit of an insurance claim. o Challenge: Non-standardized claims created a data integration problem. o Selected a binary classifier logistic regression model as the best predictor. o Defined a new workflow for metadata and implemented a process upstream to standardize fields. • Gained experience in the software side and the ML side of building production scale AI applications, using distributed Apache Spark frameworks, to score very high data volumes. This is particularly important for those new to data science. 30
  • 29.
    WHAT WAS PRODUCED? •Developed a tool that tracks customer behaviors and purchasing at a national grocery chain. o Designed and created a data pipeline that included geospatial data for a broader understanding of the entire region. o Enabled each business division to analyze and quickly respond to customer needs due to the evolving changes in supply chains.What did you create?Was it a single tool, a portfolio of services, or an analytic platform? 31 If you work in industries that have been significantly impacted by the Coronavirus, now is the time to think about what analytic products can meet these evolving needs.
  • 30.
    WHO GOT VALUEAND HOW? • Advise undergraduate and graduate students, mentoring through the process of determining a career path by channeling their commitment and passion for research and analysis. o Leveraged a student's interest in the local business environment by coauthoring a paper on "Startups in Los Angeles.” o Resulted in the student getting a Research Assistant job at the L.A. County Offices. o Received the 2019 Outstanding Graduate Student Research and Mentorship Award at my college based on her nomination. Think about who got value from the success of an effort or project. Demonstrate and quantify results here. 32
  • 31.
    RESUME SUMMARIES A DataEngineer and a Database Engineer with extensive project management experience as a Certified PMP, creates major data-driven value opportunities for the company. • Turns around troubled projects and reenergizes the team to regain commitment. • As an experienced financial analyst, uses data storytelling techniques to create informative graphs with comprehensive narratives. • Works collaboratively throughout the analytics lifecycle and has particular expertise in communicating and operationalizing the results. In resume summaries it is also important to think about the issues that we’ve covered. 33
  • 32.
  • 33.
    LESSONS IN RESUMEWRITING • Follow best practices • Know which role you are aiming for • Tell a coherent story with what, why, and how components • Reach out to people 35