bias

Bias in Neural Nets

Image: http://www.buildhealthnaturally.com/care-and-feeding-of-your-brain/

Large companies

Bachelor's
degree

Non-traditional route to
data science

Anecdata, Observations, and Advice

20% of my career officially employed as a Data Scientist

#IndividualContributor4Life

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

A Short History of My Own Career

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

Student Systems Administrator

Systems Engineer

Proposal Writer

Developer

Data Scientist

job*n = career

What is a career, really?

contribute to society

personal satisfaction

intellectual growth

practical needs

mental and emotional needs

food

shelter

saving for retirement

The purpose of a career is to find our global maximum.

Jobs are our gradient... ascent.

A career is a continuous calibration.

A job is one point in that journey.

Jobs can teach you what you like...

and also what you don't like.

It is safe to assume that you will have many jobs over the course of your career.

It takes two years to get comfortably acclimated in tech job.

2

spring

summer

autumn

winter

https://www.cmurrayenterprises.com

Seasons of Change

Seasons of Change originally from Cynthia Murray

and on the subject of changing jobs...

Changing jobs is completely ok.

Strategize your career.

Make each move for a reason.

Run to an opportunity, not away from a reality.

Don't be afraid of lateral moves or steps backward in pay.

Carefully taken, these choices can superpower your career.

$$$

Student Sysadmin

Proposal Writer

Systems Engineer

Developer

Data Scientist

First entry level job: omg real jobs pay money!

Promotion!

Learn to write and communicate.

Learn to engineer production software.

Learn to build and deploy models.

Field Change:
Senior to entry level.

Field Change:
Lateral move.

the day to day of tech jobs

Money makes the world go round.

You will be hired to improve a company's bottom line.

Replace human effort with automation.

Create new products to increase revenue.

Save Money

Make Money

$

You will be a change agent.

You have to build trust with the business and your users.

It is unlikely that you will be working on the cutting edge of technology, constantly creating new breakthroughs.

(Sorry).

You will build things that
make a difference.

You will implement solutions that
help others.

However...

Be prepared to do a lot of software engineering as a "Data Scientist."

Stack Overflow is there for all your syntax needs.

Good software design, not so much.

Building a model

Getting a model into production

Your Time

"ModelOps"

Design with the entire model lifecycle in mind.

Image from SAS

Getting suitable data training data is usually the biggest challenge.

Trust, but verify.

There will be meetings. So. Many. Meetings.

post-covid

pre-covid

Carve out and protect your deep work time.

useful tools and workflows

Reproducibility.

Do future you a solid.

1

2

Your tools and workflows should support two things.

Code

Data

Environment

Reproducibility is the cornerstone of data science.

notebook

source control

code

env

data

Image: jupyter.org

code

env

data

Learn to use git.

These commands will solve 95% of your needs:

git clone REPO_URL

git checkout BRANCH_NAME

git checkout -b BRANCH_NAME

git push -u origin BRANCH_NAME

git pull

git commit -a -m "commit message"

git push

git checkout main
git pull
git checkout BRANCH_NAME
git merge main
git push

Push a new local branch to remote.

Create and checkout a new branch.

Commit all unstaged changes.

Pull down latest main changes, merge changes into your branch, and push.

code

env

data

Regexes are a superpower.

compliments of xkcd.com

code

env

data

package management

environment management

poetry

anaconda

virtualenv

code

env

data

Beware of pickle (and pandas pickles) for long-term storage and serialization.

Use where clauses.

A couple general tips:

Source control your queries.

Data management approaches for data science efforts will be highly dependent on your company datastores and platforms.

Bottom Line

Make it easy for other people to:

,

use

your work.

and

reproduce

integrate

,

Goal for present day you:
Do your future self a solid.

Take the time NOW to write your docstrings.

Figure out a personal filing system approach...

Digital clutter is inevitable.

You are going to create, receive, and store A LOT of files.

...and stick to it.

Make sure everything on your work machine is backed up.

Use company cloud and network storage from Day 1.

advice for career success

Strive for excellence in your work.

"...governed by good ideas and executed with superb craft."

- Edward Tufte

Good communication skills are essential.

Advocate for yourself.

Negotiate your salary, and ask for raises.

Keep records of your successes,
and learn when and how to promote them.

It's not just who you know, ...but it's a lot who you know.

miscellaneous, unsolicited advice for personal success

Always Be Learning.

Take a Personal Finance class.

If you can't, spend some time learning the basics of personal finance before you graduate.

(COMM 2730)

Pay yourself first.

Meet your employer 401k match: it is free money.

Start saving early and consistently.

Income is what you earn, wealth is what you keep.

$

Personal Finance CliffsNotes:

Invest in yourself (i.e. save for retirement).

Compound interest is what will make you rich.

It is completely ok if your job is not your passion.

At the end of the day, a job really is just... a job.

There are always tradeoffs.

Your career is only one component of your life.

Remember, a career - and life - is a journey.

Stay open to experience along the way.

and Opinionated Advice on

Building a Career in Tech

Rachel House

Practical Tips for

bias

A Short History of My Own Career

job*n = career

2

Seasons of Change

the day to day of tech jobs

$

$

useful tools and workflows

1

2

advice for career success

miscellaneous, unsolicited advice for personal success

Always Be Learning.

$

$

$

and, above all...

thank you!