Greg Detre - Making Data Mistakes

How to kick off a successful project

"Sorry, there's no room on our roadmap to help with that this year." I remember hearing those words from the IT team with a sinking feeling. We had nearly finished the algorithmic work for our latest data science project, and now we needed IT's help to get it into production.

Collaboration

How to make your software team much more effective

Programming is the closest we have to magic in the real world. Just like magic, you can conjure rich and powerful creations with the power of your own mind, given the right tools and training. Just as powerful magic could lay waste to an entire army, powerful programming can do

Collaboration

Postmortems: The single most important keystone habit for a team (step-by-step guide)

I’m an ill-disciplined person. My habit of reading books about habit-formation is stronger than any of the habits they’ve help me form. Perhaps because of this, I’m interested in the idea of keystone habits [https://callmebetty.com/17-keystone-habits-what-they-are-and-how-you-can-use-them-to-skyrocket-your-success/#:~:text=Charles%20Duhigg%20defines%20a%20Keystone,of%20hacking%20your%

Data Science

Make it work, make it right, make it fast

I first heard this phrase from Gael Varoquax [http://gael-varoquaux.info/] (one of the core scikit-learn committers) in a talk[1], and I loved the super-concentrated wisdom of it: * Make it work. First of all, get it to compile, get it to run, make sure it spits out roughly the

Data Science

How Humans and Machine Learning Deliver the Best Result (Podcast)

I was recently honoured to be a guest on The GrayMeta Podcast - Metadata Matters. In this podcast, I have the opportunity to discuss my perspective on Metadata and how to structure human teams to effectively work with data generated by machine learning. Hosted by the incredibly talented Matt Eaton.

Collaboration

How to reproduce your analysis

But first, an exciting announcement! I'm currently hiring for two Machine Learning Engineering roles, both doing interesting work with high-energy startups led by great people on good missions. Scroll to the bottom of this article for more details! -------------------------------------------------------------------------------- How to reproduce your analysis > “But didn’t you say you

Collaboration

How to run an Idea Stampede to make a group gallop

aka How to have a conversation productively with 10 people at once (aka How to run brainstorms like a pro)

Classification

What is precision vs recall, and why should I care?

What do these tasks all have in common? * Looking at an X-ray and deciding whether someone’s leg is broken * Screening a CV for suitability for a job post * Identifying faces of terrorists in a crowd They are all examples of classification tasks, where we need to make a binary

Data Science

How to Cheat at Data Science (with help from Centaurs and the Wizard of Oz)

What does it take to run a high-performing data science team? In this 15-minute video, I talk about two tools that I’ve found invaluable for doing just that: * Wizard-of-ozzing is a technique for product discovery that can be very helpful for evaluating potential problems to work on [6:00]

Management

Why we're bad at estimating (and what to do about it)

A product manager recently asked me if one can get better at knowing whether things are easy or hard, and how long they will take. The good news is that with practice, you can help people estimate much better than they can estimate on their own. Here's how. Understand the

Software Engineering

A little functional programming goes a long way

Object-oriented and functional programming styles have both been around for a while, and have their own committed adherents. It’s almost certainly possible to be effective with either. But on balance, you should prefer functional programming.

Data Science

What does it take to transition from Senior to Chief Data Scientist?

I’ve been asked this question by a few highly competent Senior Data Scientists over the years. Even just asking the question is usually a good sign that they have a sense of the answer - a sense that the skills that make you a good Senior Data Scientist may

Data Science

What makes a great data scientist and data science team? (Podcast)

I was recently honoured to be a guest on The Data Pubcast [https://anchor.fm/data-pubcast/episodes/The-Data-Pubcast--Pint-3-ek3l21/a-a3b42vj] - a podcast about making data accessible to everyone, hosted by the incredibly talented Nick Latocha and Andy Crossley. In the episode [https://anchor.fm/data-pubcast/episodes/The-Data-Pubcast--Pint-3-ek3l21/a-a3b42vj] we discuss:

Data Science

What is Machine Learning?

Machine learning algorithms are faced with the same challenge you had as a pupil at school. There’s going to be an exam. The higher you score the better, and you'll be given some exams from the past with answer sheets to study from.

Data Science

What is Overfitting?

Overfitting is a common problem in Machine Learning. But what does it actually mean? And why is it an important problem for Data Scientists to overcome?

Programming

How to write good software faster (we spend 90% of our time debugging)

If we spend the majority of our programming time and effort on debugging, we should focus our efforts on speeding up our debugging (rather than trying to write code faster).

Data Science

How to know if your recommendations algorithm is actually doing a good job

I led the team that built Channel 4’s recommender system for All 4 in 2016. It started out as a straightforward project. But after getting lost in a rabbit hole trying to devise a score for ‘provocativeness’ and ‘serendipity’, I learned the single most important lesson about data science.

Philosophy

Necessary vs Sufficient

Philosophers and mathematicians make a distinction between ‘necessary’ and ‘sufficient’. Air is necessary for human life, but it’s not sufficient. In other words, if you don’t have air then it’s game over. But if you do have air, you still need some other stuff to survive. So

Management

How to make your data science team faster (and speed up progress)

What do you do if your CEO complains that progress is too slow? While it’s tempting to think you just need to buckle down and “get more done”, more often than not, simply trying to "increase throughput" won’t work, especially if you’ve misdiagnosed the root cause of the problem.

Software

Why are my emails being sent to spam - A guide for startups using GSuite

“This email to an investor was marked as spam! They only noticed it by accident. Why is this happening?”

Software

The Difference: Throughput vs Latency

Latency and Throughput are important concepts for data scientists. How are they distinct? And why is this distinction important, not just for technical systems, but also for team performance?

Data Science

Will data scientists become obsolete in the next 10 years?

Will tech companies still need data scientists, or are they destined to become just another blip in the history books – disappearing from the world like film projectionists and bowling alley pinsetters before them?

Job-hunting

How to find a job through a recruiter

‌‌Every one of the dozen or so job offers I’ve received in London in the last 10 years has been through a friend’s recommendation, or through recruiters. If you have a friend who works for the company or knows someone senior, a strong recommendation is the best entrance

Management

How to make good, quick, collective decisions

This is a recipe for making good, quick, collective decisions: * Write a proposal with: a clear statement of the goals you're trying to achieve (always start with the goals!), your suggested approach, and maybe an argument for why it's a good idea. (A Google Doc is good for this kind

Software

Technical debt isn't like financial debt

I hope you weren’t hoping for a blanket answer to the question of exactly when to pay off technical debt. If you were, the answer is Tuesday. Almost all interesting decisions are about risks and trade-offs. This is a framework for making our intuitions about technical debt more rigorous.