An Introduction to Glassdoor Engineering

Glassdoor was founded in 2008 with the goal of driving workplace transparency through our mission of helping people everywhere find a job and company they love. Glassdoor, through its many features, helps both our users and our employers understand the areas of their work environment that need attention and in turn pushes for a healthy and employee friendly work environment so we can all bring our 100% every day. To drive our mission home, Glassdoor engineering team works on solving complex problems at scale using rich data. By rolling out our engineering blog, we plan to share some of our learnings with the broader tech community.

Culture

The highlight of our engineering culture is that it thrives on collaboration, learning, and empowerment. Each member of our team has a part to play in driving innovation forward and an opportunity to influence product direction and technology. We believe that exchanging ideas helps produce the best solution. In the process, we learn from each other and get a broader perspective on architecture and design. Every 6 months in our 3-day engineering hack-a-thons we wear multiple hats and work at a lightning speed to push our ideas forward. Many of our winning ideas have become part of the Glassdoor product and tech stack, which is incredibly empowering.

People

Working at a company with a mission of helping people at its core is the power and passion that drives us. As a diverse team of engineers, we all bring our unique perspectives to the technology and product conversation as we work to make a difference through our product. We hold a high bar for hiring talented people and invest in developing and empowering them with large decision-making scopes. We value the success of the team as a key part of personal success.

Technology

As a platform trusted and loved for sharing authentic information on companies and surfacing core insights by processing millions of pieces of content from our user community, engineering is at the core of driving this scale and quality for -

Our 50¹ million unique monthly users around the world.

Surfacing high quality content & insights on 1 million employers in 190 countries².

Serving 10 million jobs and more than 60 million pieces of content in the form of reviews, salaries, interviews, benefits and other workplace insights³.

Multiple engineering disciplines come together to serve our users, each tackling unique challenges with scale, quality & performance, constantly striving for engineering and product efficiency.

Our applications are built using React, EmotionJS, Sass bundled using Webpack, and Spring Boot for both Java and Kotlin-based services. UI rendering services are built on NodeJS/Express and aggregation layers use GraphQL and Apollo Federated Graph. We always look to integrate with the latest browser performance features such as ServiceWorkers, Loadables & WebWorkers wherever applicable to understand and improve our page performance.

Glassdoor’s mobile apps are fully native and built on Swift and Kotlin. We are also looking at Kotlin multi Platform to see if there are opportunities for sharing code between iOS and Android platforms. Also, our mobile team is currently experimenting with a unique idea of collections to help our heavy mobile users to gather relevant information in one place for ease of navigation.

We use Solr for our jobs search and recently migrated our content search to Elasticsearch. The backend infrastructure relies heavily on Kafka, Storm, Cassandra, and Redis clusters along with event streaming infrastructure using Kinesis, Lambda, and Firehose under the hood.

Our machine learning scientists and engineers are focused on constantly innovating to surface key insights from reviews, provide salary estimates, recommend personalized jobs, moderate user-generated content, and recommend relevant pieces of information to our users. Members of ML teams are also embedded within product teams based on the project need and team structure. These teams utilize Python, Java, Presto, Hive, Spark, Airflow, Cassandra, Redis, Apache OpenNLP, etc. to build end-to-end machine learning solutions to power Glassdoor products by the intelligence from our data and algorithms. Data scientists also work closely with our data engineers to drive data-driven decisions as they both leverage Kinesis, Lambda, Spark, Zeppelin, etc. We use Qubole to scale and provision on-demand infrastructure to efficiently manage our clusters. The team relies on open source technologies like Apache Hadoop for data processing, Airflow for jobs orchestration, Singer.io for API source ingestion, Grafana for ops monitoring, dbt (data build tool) to transform, etc.

Core engineering and dev ops teams work horizontally to solve for network, security, reliability, deployment efficiency, and support AWS cloud and infrastructure management. These teams build the backbone of Glassdoor engineering, ensuring we all operate as a high-functioning team of engineers. We are starting to invest heavily in kubernetes, containerization & serverless technologies.

Our engineering excellence principles and values help drive our technology vision and best practices. To effectively drive the changes involving cross-team efforts, we have organized ourselves around functional skills, as Guilds. Initiatives driving our engineering excellence and Guild goals create leadership opportunities for engineers outside their respective teams to make a larger impact.

It is such a humbling and fulfilling opportunity to help millions of people make one of life’s most important decisions — finding a job. While Glassdoor has its presence in many countries, Glassdoor engineering is located primarily in San Francisco (CA), Mill Valley (CA), and Chicago (IL). We also have remote members spread across the country and are always looking for more stellar talent to join us. As our teams continue to tackle tech challenges, we will share our learnings with our broad tech community through our engineering blog.

[1] Source: Google Analytics, Visits represents peak monthly visits in CQ2'20 [2, 3] Source: Glassdoor Internal Data, March 2020