February we're featuring a Job Fair for the first time. CBRE (http://www.cbredev.com/careers/) and OfferUp (https://offerupnow.com/jobs/) are participating companies. Brian Weber a SRE at Twitter will share a talk on unit testing. Jed Dougherty of data iku will share a talk on deploying Python predictive models as APIs.
5:00 PM Job Fair with CBRE and OfferUp (starting b4 talks)
6:00 PM Talk format event opens for mingling
6:35 PM First speaker
7:05 PM Break
7:30 PM 2nd Speaker
8:00 PM Lightning Talks
8:30 PM Move to J&M
Speaker: Brian Weber (https://www.linkedin.com/in/brian-weber-2423b55?trk=hp-identity-name)
Brian Weber is visiting from the Bay Area. He's been developing primarily in Python for five years.
He has mostly worked with monitoring/alerting services, deployment tooling, and thrift APIs. He's written and maintained services of many sizes.
He has worked at Twitter, Pinterest, and Facebook. Unit testing is near and dear to his heart. In his free time, he skis, cooks, cycles, builds home networks, and kayaks.
Mocks for Unit Testing
Is it really necessary to call an external API when running tests? How can I tell if the failure was on my code I'm importing? How do I prevent global state changes when running a test?
Unit and acceptance testing is important in any software development. Python provides a set of tools for unit testing. The mock library is an important tool because it allows you to test your code without actually affecting anything outside of your code.
This talk will be centered around various examples of code examples to demonstrate the following:
* What code should be mocked for a successful test, and how to examine your code to decide
* How to write basic tests with mocks
* How to leverage the internals of the mock library to write successful tests
Slides for the talk in RST format: https://github.com/mistermocha/python-mock-talk/blob/master/mocktalk.rst
Speaker: Jed Dougherty (https://www.linkedin.com/in/jediv/)
Jed Dougherty is a Data Scientist working to build the world’s best collaborative Data Science platform at Dataiku. Before coming to Dataiku he worked in the fields of event detection, recommendation systems, and survival analysis in the fields of breaking news and child welfare.
Deploying Python predictive models as APIs
While constructing data pipelines and building models is a core part of the Data Scientist’s job, an often-forgotten facet of the toolkit is how to actually move models into production. In this course we will build a simple model for predicting spam in hotel reviews using Python. We’ll then take that model and expose it as an API using several different tools.
This session covers various deployment strategies for serving a python machine-learning model as an API.
Many business applications can make good use of real-time scoring using machine learning, and one of the most approachable and easy languages to use to build these models is Python.
The goal is to show the audience how to actually take a trained python model and turn it into an API. We’ll start very simple and cover increasingly complex deployment strategies. Throughout, we will consider the API throughput and resource tradeoffs, and benchmark our solutions.
We have Brett Vitaz and Michael Patterson joining with lightning talks.
CBRE (http://www.cbredev.com/careers/) sponsoring food for our meeting.
data iku (http://www.dataiku.com/) sponsoring beverages for our meeting.