ISSUE 379  ·   March 22, 2022

In this week’s issue, Conor Dewey joins Data Elixir with an eye towards data teams and analytics. Conor is a former data scientist at Squarespace and is currently a product manager at Metabase, an open source approach to business intelligence.

Trends

The 2022 AI Index Report

The latest edition of Stanford's AI Index Report considers data from a wide range of organizations and offers insights into the current state of AI and where things are going. Covers technical capabilities, research and development, ethics, AI policy and governance, and more. 
Stanford University | Human-Centered AI

Organizations

Organizing and Scaling an Effective Data Team

I quite like this post on how to scale your organization’s inaugural data team. I’ve spent more time around earlier-stage companies the last few years so the emphasis on the 1-4 member stage stood out to me as particularly useful. Those early days can be a difficult trek.
Rob Dearborn

Founding an Analytics Engineering Team

Staying on the team-building theme, this write-up from dbt does a good job of outlining a similar journey through the lens of analytics engineering. The “State of Analytics Before AE” and “Justifying & Starting the AE Team” breakdown is a great way to look at things in my experience.
dbt Blog | Nate Sooter

What I’ve Learned About Documentation for Data Teams

With enough team growth, it’s inevitable that documentation becomes a foundational topic of discussion for data teams. It’s a hard skill to master individually and even harder as a team. Brittany Bennett shares her experience of “taking her team dark” for a month and overhauling how they thought about data documentation.
Brittany Bennett

Sponsored Link

Understanding and Overcoming Four Types of Biases in AI

There are four types of biases found in machine learning models. These are algorithmic bias, sample bias, prejudicial bias, and measurement bias. How do each of these biases arise and how are each of them mitigated? Read this article to understand how you can produce better business outcomes by training AI models to do precisely what they are meant to do.

Tutorials, Projects & Opinions

Imbalance Detection for Healthier Experimentation

This post from the experimentation team at Etsy is a good one. If you’re frequently running A/B tests with complex required instrumentation, things go wrong from time to time. The question is: how do you know when you’re getting balanced, representative results and when things are imbalanced? 
Etsy - Code as Craft

Is Facebook Prophet suited for doing good predictions in a real-world project?

This real-world look at Prophet explores its strengths and weaknesses to help you determine if Prophet is a good fit your own forecasting projects. Covers feature engineering and modeling, interpretability, and maintenance, with tips and tricks along the way.
Jules Bertrand & Hugo Vasselin

Data Science Project Quick-Start

Starting out on a new data science project can be a bit overwhelming. The surface area is often vast with a number of directions you could go. One way to make your life easier is to have a framework. In this post, Eugene Yan lays out exactly that, from understanding the problem to defining requirements to ultimately digging into the data.
Eugene Yan

Deploy a Data Stack 3x Faster w/o Data Eng

Modern Treasury did it — see how your team can do the same
// sponsored

Discussions - Click through to read or participate

How did you like this issue of Data Elixir?

👎    1     2     3     4     5   👍

Was this email forwarded to you?  Sign up here >>