Skip to content

The Mux Blog

We're a team of engineers, marketers, designers, all passionate about video and the work we create together. Welcome to our blog about video.

Search
Quantifying packaging overhead

Published on May 1, 2019 (almost 6 years ago)

Quantifying packaging overhead

Matthew Szatmary
By Matthew Szatmary7 min readEngineering

Mux makes adding video to your app or website as easy as making a single API call. But behind the scenes is a large multistep process to analyze and transform the video into something that can be easil...

How we improved Tensorflow Serving performance by over 70%

Published on February 26, 2019 (almost 6 years ago)

How we improved Tensorflow Serving performance by over 70%

Masroor Hasan
By Masroor Hasan13 min readEngineering

In this blog, we’ll focus on techniques that improve latency by optimizing both the prediction server and client.

An adventure in OpenAPI V3 code generation

Published on February 20, 2019 (about 6 years ago)

An adventure in OpenAPI V3 code generation

Phil Cluff
By Phil Cluff26 min readEngineering

Learn about Mux's journey to support 10+ languages with SDKs, leveraging OpenAPI V3, and tackling code gen challenges.

Multi-CDN support in Mux Video for improved performance and reliability

Published on February 12, 2019 (about 6 years ago)

Multi-CDN support in Mux Video for improved performance and reliability

Scott Kidder
By Scott Kidder10 min readEngineering

See why you should care about CDNs and how Mux Video automatically leverages multiple CDNs to deliver the best network performance and viewing experience.

Video Access Log Processing with Apache Flink

Published on October 24, 2018 (over 6 years ago)

Video Access Log Processing with Apache Flink

Scott Kidder
By Scott Kidder9 min readEngineering

This is an overview of the stream-processing system built by Mux to handle large volumes of video-streaming logs with Apache Flink and Kafka.

Real-time dashboards powered by Elixir

Published on October 8, 2018 (over 6 years ago)

Real-time dashboards powered by Elixir

Jacqui Manzi
By Jacqui Manzi14 min readEngineering

Using Erlang’s OTP and Phoenix’s Channel / Presence to manage large real-time data sets across many channel topics Our real-time analytics product requires a scalable system with high availability a...

Published on January 12, 2018 (about 7 years ago)

How We Used Machine Learning to win at HQ Trivia

Ben Dodson
By Ben Dodson8 min readEngineering

For the past three months, half of us at Mux have been obsessed with HQ Trivia, while the other half silently regretted their Android purchases. When that HQ push notification hits at 12, the iPhones s...

Using Percentiles To Identify Outliers and the Typical Experience

Published on December 22, 2017 (about 7 years ago)

Using Percentiles To Identify Outliers and the Typical Experience

Scott Kidder
By Scott Kidder6 min readEngineering

The most overused measurement is the arithmetic mean (average). In nearly all cases you're better served by using percentiles instead of averages.

Using Control Streams to Manage Apache Flink Applications

Published on March 7, 2017 (almost 8 years ago)

Using Control Streams to Manage Apache Flink Applications

Scott Kidder
By Scott Kidder6 min readEngineering

Learn how Mux has used control streams to manage Apache Flink applications with zero-downtime.

Discovering Anomalies in Real-Time with Apache Flink

Published on February 15, 2017 (about 8 years ago)

Discovering Anomalies in Real-Time with Apache Flink

Scott Kidder
By Scott Kidder7 min readEngineering

Mux receives a continuous stream of video-view event data, every second of every day. Some of these views indicate errors during the delivery or playback of video. How can we detect error spikes in rea...