Deputy Builds Intelligent-Scheduling App on AWS

Deputy’s mission is to improve the lives of shift workers, who account for some 60 percent of the global workforce. Its mobile application is unique in its use of sentiment analysis, a feature built with machine learning (ML), to influence scheduling decisions. Deputy is then able to quickly understand how to optimize an individual’s schedule to enable them to bring the best version of themselves to work each day.

The business is also focused on making work life easy for business owners and employees. By empowering staff to be stakeholders in the scheduling process, workplace engagement increases.

Ahmed and Shelley believe technology—especially cloud computing—is an enabler that drives significant outcomes for their customers. In its decade-long existence, Deputy has evolved its offerings as AWS has introduced new services, and it continuously looks at ways to improve efficiency. One of the aspects it values most about cloud computing is the agility offered for experimentation. “You can always undo something with AWS,” Ahmed explains.

For storing vast amounts of customer data, Deputy uses Amazon Aurora with MySQL. The startup also uses Amazon Redshift as a data warehouse for internal company data and analytics, and Amazon Kinesis Data Firehose to load streaming data into Amazon Redshift. A couple of years ago, Deputy launched a facial-analysis feature with biometric validation for employees to clock in and out, built using Amazon Rekognition.

The ISV has also been evolving its architecture to increasingly rely on serverless infrastructure as code. Everything the company has done in the past two years is based on AWS Lambda, and the team is moving toward a completely serverless future.

“I’m excited about Lambda because of the scalability it provides. It’s a secure way of delivering our application to end users and allows for simple testing and deployment,” Ahmed says. He believes the shift to serverless computing is a fundamental value creator because it saves so much time and money by deploying infrastructure on demand. In a recent example, Deputy launched an Employee Onboarding feature built entirely on Lambda that reduces the time for customers to onboard a new employee from 30 minutes to just 1 minute. This adds up to significant savings for customers in the retail and hospitality industries, which typically have high employee churn rates.

Since its inception, Deputy has grown 100 percent or more year-on-year, bringing its scheduling solution to new markets at an impressive rate. It operates out of AWS data centers in the US, UK, and Australia to ensure data sovereignty as well as General Data Protection Regulation (GDPR) compliance. “It has been tremendously helpful to be on AWS when we acquire new customers in the UK or US, and we can do this only because of the global footprint of AWS availability zones. There are growing pains everywhere, but our infrastructure isn’t one of them,” Ahmed shares.

The company has been able to confidently scale with AWS as its cloud foundation. Issues and minor outages have naturally occurred, but these are quickly resolved and transparently communicated. “Our experience with AWS has been that if there’s an issue, it happens only once,” Ahmed says. “I know that AWS is there to help us when we need it. The team has been very customer-focused in good and bad times.” Deputy has subscribed to AWS Business Support the past few years to get quick, reliable help whenever it faces a scaling or technical issue.

Despite an increase in its uptake of AWS Managed Services, Deputy has been able to control infrastructure costs and continually invest in innovation. “AWS is constantly improving its products and services while remaining affordable, which is similar to the philosophy we apply with our customers,” Ahmed relates.

Empowering the Deputy software to make decisions akin to how humans make them requires intense computation power and continuous adjustment to learning models. “By simply pressing a button, we can build a schedule from millions of different possible combinations. This is almost impossible for a human to do, especially in a tight time frame,” says Ahmed. Over the past few years, the software has evolved from simply being a tool for schedule creation to serving as a knowledge hub using data and ML to make managerial decisions.

Furthering its data-driven agenda is Deputy’s recent formation of a data-science team, tasked with building more AI and ML functions in-house. One of the team’s first creations is the “Late to Work” report, compiled from insights gleaned from its global data stores. This area of the business will continue to grow as Deputy trials Amazon SageMaker and eagerly awaits the launch of Amazon Forecast to expand its business.

“Deputy wouldn’t exist today if it weren’t for AWS. There’s so much opportunity ahead of us, and I couldn’t be more excited for AWS to be part of our journey as a partner,” Ahmed concludes.