Profile picture
, 7 tweets, 2 min read
To learn how to design machine learning systems, I find it really helpful to read case studies to see how great teams deal with different deployment requirements and constraints. Here are some of my favorite case studies.
Topics covered: lifetime value, ML project workflow, feature engineering, model selection, prototyping, moving prototypes to production. It's completed with lessons learned and looking ahead!

medium.com/airbnb-enginee…
Netflix streams to over 117M members worldwide, half of those living outside the US. The company uses machine learning to predict the network quality, detect device anomaly, handle predictive caching.
medium.com/netflix-techbl…
This article explores the tradeoff between complexity and interpretability, performance and ease for production, for the classic task of fraud detection. (4/n)
eng.lyft.com/from-shallow-t…
Instacart uses machine learning to solve the task of path optimization. The article explains the entire process of system design, from framing the problem, collecting data, algorithm & metric selection, with visualization thrown in. (5/n)
tech.instacart.com/space-time-and…
Fashion Taste API attempted to qualify people's fashion taste using machine learning! (6/n) medium.com/hackernoon/how…
This article walks you step by step through a canonical example of the ranking and recommendation problem.

Four main steps: system design, personalization, online scoring, and business aspect (7/n)

medium.com/airbnb-enginee…
Missing some Tweet in this thread? You can try to force a refresh.

Enjoying this thread?

Keep Current with Chip Huyen

Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

More from @chipro see all

Profile picture
Chip Huyen
@chipro
To better understand the technical hiring pipelines, I analyzed 15,897 interview reviews for 27 major tech companies on Glassdoor. I focused on interviews for software engineering related roles, both junior and senior levels. These are some of the main findings. (1/n)
Each review consists of:

- result (no offer/accept offer/decline offer)
- difficulty (easy/medium/hard)
- experience (positive/neutral/negative)
- review (application/process/questions)

The largest SWE employers are Google, Amazon, Facebook, and Microsoft.
Strong correlation bw onsite-to-offer rate and offer yield rate (% of candidates who accept their offers). The more selective the company is, the less likely a candidate is to accept their offer. Candidates that pass interviews at FAANG are likely to have other attractive offers.
Read 13 tweets
Profile picture
Chip Huyen
@chipro
This thread is a combination of 10 free online courses on machine learning that I find the most helpful. They should be taken in order.
1. Probability and Statistics by Stanford Online
This self-paced course covers basic concepts in probability and statistics spanning over four fundamental aspects of machine learning: exploratory data analysis, producing data, probability, and inference.
online.stanford.edu/courses/gse-yp…
2. Linear Algebra by MIT

Hands down the best linear algebra course I’ve seen, taught by the legendary professor Gilbert Strang.
ocw.mit.edu/courses/mathem…
Read 11 tweets
Profile picture
Chip Huyen
@chipro
I'm working on a book on machine learning interviews so I've been spending the last few months talking to companies about their hiring process for ML roles. This thread is a summary of what I've learned. It will be updated as the book progresses. (1/n)
The average interviewer gets very little training. You start your full-time job. You shadow a few interviews. Then you're on your own. As a result, interviews are wildly different even within the same company.
I ask interviewers to give me their most frequently asked questions, and show them questions that other interviewers ask. I notice the pattern that if an interviewer doesn't know the answer to a question, they immediately flag it down as bad.
Read 18 tweets

Related threads

Profile picture
John Cutler
@johncutlefish
I was speaking with an executive with 20+ years of experience recently. And I asked them if they had ever seen a feature/effort “flop”. And they said No. And I found this very curious. So I dug... (1/6) #design #prodmgmt #agile #devops
...in many cases, they claimed “poor execution” had left the “on their back foot”. Meanwhile, new silver bullets had emerged. So they didn’t really know — or weren’t focused on — whether the original idea worked or not (2/6) #design #prodmgmt #agile #devops
...in other cases, they described competitive pressures or “market shifts” that again left them “late to the party”. Notice the trend: it is all about speed and time in their mind. Perfect ideas that couldn’t be executed on quickly enough. (3/6) #design #prodmgmt #agile #devops
Read 6 tweets
Profile picture
Himanshu Khanna
@SparklinGuy
In Dec 2018, @WeAreSparklin introduced a new series - One thing at a time.

Think of ‘OTAT’ as a bite-sized modification to a platform, to heighten the user experience.

We have written fortnightly about 8 different products including Uber, Whatsapp, Zomato

A thread👇🏼
#design
One easy to way to skip the thread and read about One thing at a time in better detail is to review our post

medium.com/@sparklin/one-…

If you’re interested in quicker learning, let me begin with the very first OTAT product improvement 🚕
Uber

Search is usually found at the top

A desktop vs a smartphone have different dimensions. 50% users prefer one-hand approach to mobile.

Why must we move forward with a search-on-top approach on mobile?

What if we moved Uber’s search to lower 40% of the screen?

Next 🍔
Read 11 tweets
Profile picture
Tusk
@TuskCulture
This is happening now. If you're interested in heritage collecting around experiences of terror and conflict, there's a Zoom presentation going on with people who have worked in this area in other countries.
Hearing first from @kostis43 about his work in the wake of the Manchester terrorist attacks of May 2017 at an Ariana Grande concert. 23 people lost their lives, most of whom were young people and their parents. Kostas is telling us about the resulting archive of the memorials.
Archive is titled "Manchester together", coming from the #StrongerTogether slogan that the city came up with in response to the attacks.
Read 30 tweets
Profile picture
Arthur Charpentier ❄
@freakonometrics
"the authors find that their algorithm can classify criminal faces with 90% accuracy … algorithm finds that criminals have shorter distances between the inner corners of the eyes, smaller angles between the nose and the corners of the mouth and higher curvature to the upper lip"
"Why would this possibly be? There’s a glaringly obvious explanation... As one smiles, the corners of the mouth spread out and the upper lip straightens. Try it yourself in the mirror" (composite faces for criminals (left) and non-criminals (right))
from "Criminal machine learning" callingbullshit.org/case_studies/c… (thanks @tomroud) - including several case studies callingbullshit.org/case_studies.h…
Read 4 tweets
Profile picture
Susanne
@susannehusebo
I’ve worked with lots of agile #teams, and some of the symptoms of a well functioning high-perfoming teams that I see are:

👇
They start their daily standups on time.
They tend to laugh a lot and have fun.
Read 86 tweets
Profile picture
Abhishek Murarka 🐂💹
@abhymurarka
💪A thread on amazing Twitter threads 😍👇
In the process of revamping my #Twitter profile (dormant earlier), I have come to appreciate the wonderful underutilized platform it is!
Some threads contain crisp knowledge capsules that can help many. Starting 1 on such curated content
1. I largely rely on this: threadreaderapp.com to find content
2. What I wish to share is what coincides largely with my interests- #Learning, #Investing, #Behavioralpsychology (#mentalmodels, #bias, etc), #Businessbooks, Industry insights/case studies, Life experiences, etc
3/n
 While it will be difficult to structure the order of content here, I will try to block threads by category (as noted above & more). Feel free to comment with any good thread that you would have come across. Cumulative learning always helps.
#vicariouslearning
Read 37 tweets

Trending hashtags

#fsi18 #correction #mentalmodels #ux #WarrenBuffet #Learning #dayfour #Fundraising101 #bigdata #WordCamp #bigtrain #bookmarked #Masterminds #tricklebrand #ciencianobr #connections #bullsandbears #agile #Haskell #AssessBetter #TheYearofInfrastructure #doh #OSS #inception #policy
Follow Us on Twitter!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!