1 of 26

Build Your First ML Pipeline

MLOPs.community + Girls in Tech

Wifi: 42Berlin_Guest Password: metaverse_guest!

2 of 26

Agenda

  1. 6:00 pm - Arrive & Pizza
  2. 6:30 pm - Introduction MLOps and GiT
  3. 6:45 pm - Workshop Introduction
  4. 7:30 pm - Break + setting up environment
  5. 7:45 pm - Workshop
  6. 9:45 pm - Wrap up
  7. 10.00 pm - Networking

Goal: Get hands on, have something you can add to your portfolio; connect!

3 of 26

Access slides

4 of 26

Intro of our Communities

5 of 26

WELCOME TO THE

Part of 20K members

around the world!

6 of 26

JOIN OUR

COMMUNITY

ON SLACK!

Click here >>>> mlops.community <<<<

7 of 26

JOIN THE BERLIN ORGANISERS TEAM!

Contact Morena for more info :)

www.linkedin.com/in/morena-bastiaansen-225b6518

8 of 26

/

9 of 26

Girls in Tech’s Mission

We are a global non-profit organisation focused on eliminating the gender-gap in tech.

We help women to succeed in tech by providing:

  • An inclusive and supportive network
  • Educational resources
  • Career development and mentorship opportunities

9

10 of 26

54

37

CHAPTERS

COUNTRIES

10

130K

MEMBERS

11 of 26

About GiT Germany

Non-profit organisation that started in 2021

Run exclusively by volunteers

2 - 4 events per month

Part of Girls in Tech Global

11

12 of 26

Our Audience - women age 25 - 45

12

Follower demographics, top 3:

Industry:

  1. IT Services and IT Consulting (11%)
  2. Software development (11%)
  3. Technology, Information and Internet (5%)

Seniority:

  1. Entry (33%)
  2. Senior (31%
  3. Director (5%)

Job Function

  1. Engineering (16%)
  2. Designer (6%)
  3. Information Technology (6%)

13 of 26

Our events

  • Educational programs
  • Career, recruitment and networking
  • Mentorship and showcasing role models
  • Social meetups

/

14 of 26

Hackathon

14

/

/

15 of 26

Recruiting Event

15

/

/

16 of 26

Our partners

We established exciting collaborations with like-minded partners:

  • universities
  • coding schools
  • Incubators
  • tech event organizers
  • corporates and startups
  • VCs

16

17 of 26

Girls in Tech Germany - The Team

Marcela

Michaela

Leonie

Yessmine

Angel

Dimitra

Barbara

Diana

Elisa

Jane

Keerthana

Crista

Carolin

Astrid

Emilly

Pallavi

Our team consists of experienced tech professionals - engineers, marketers, consultants, designers and business developers.

Together, we are committed to our cause of closing the gender gap in tech.

Olivia

Simantini

/

18 of 26

Do you want to get involved?

18

/

Become:

  • Volunteer
  • Advisory Board Member / Project Lead
  • Ambassador
  • Mentor

/

19 of 26

Let’s keep in touch!

19

/

Barbara Vargas

barbara.vargas@girlsintech.org

Data Analyst

Girls in Tech Germany

Email us on germany@girlsintech.org

Website: https://germany.girlsintech.org/

IG: @girlsintech.germany

LinkedIn & Eventbrite: Girls in Tech Germany

/

20 of 26

Workshop Intro

  • Type of task πŸ’»
    • Text classification
    • Natural Language Processing (NLP)
    • Goal: assign a class to input text
    • Many use cases, such as
      • Sentiment analysis (customer review score prediction, comment section analysis)
      • Intent detection (user intent classification for Chatbots, search query intent)
      • Spam detection
      • Language identification
      • …
    • Today, we will be working on an intent detection task 🧐
  • Dataset πŸ‘€
    • Banking77: online banking queries annotated with their corresponding intent
  • Goal πŸ₯…
    • Build a model which can predict the intent for a previously unseen user query

21 of 26

3 parts, 10 tasks

  1. Data exploration & modeling

β†’ Fill the gaps in: notebooks/text_classification.ipynb

  • Building a ML pipeline

β†’ Fill the gaps in: notebooks/test_pipeline.ipynb

  • Building your app using Streamlit

β†’ Fill the gaps in: workshop/streamlit_ui.py

If you are stuck, we are walking around and will help you out :) 🀝

22 of 26

Workshop Intro

First, let’s make teams of 2-3 people πŸ‘―

23 of 26

Setting up the environment

  • Create a github account if you don't have one
  • Create a new repository based on this template:
    1. Go to URL github.com/mlopscommunity/build-your-first-ml-pipeline-workshop
    2. Use this template -> Create new repository
  • Open repo in codespaces

Let’s all take 5mins to set this up :) ^

24 of 26

A video demonstration of how to setup

25 of 26

Big THANK YOU to all the organisers!!!

And everyone else who helped us to make this workshop happen free of cost :)

you rock

Marina

Jean

Barbara

Stephen

Francesco

Liza

Marcela

26 of 26

Thanks for your time!