AWS Business Intelligence Blog

How Wisso embedded analytics in modern applications and cut BI costs by 80% with Amazon QuickSight

This post was co-written with Jussi Railakari from Wisso.

Wisso is a consulting and service partner specializing in cloud services, Internet of Things (IoT), and application development. They needed to help their client enhance analytics for their kitchen operation management application, which helps professional kitchens manage food safety, inventory, and related documentation. In this post, Wisso shares how using Amazon QuickSight for embedding analytics in modern applications helped to speed up dashboard development and increase user adoption in only 5 months.

The challenge

At Wisso, we’re not just a consultant—we provide ongoing support to maintain end-to-end continuity for client services. On behalf of our client, we provide a kitchen operations management software for commercial kitchens. The digital monitoring tool helps kitchens comply with food safety standards, reduce food waste and related carbon emissions, and comply with reporting requirements.

Commercial kitchen managers, from workplace cafeterias to chain restaurants to food factories that produce over 10,000 meals every day, require a kitchen operations management system to manage HACCP, a food safety system. HACCP is a comprehensive set of principles for assuring food safety by managing biological, chemical, and physical hazards from food production, from procurement to manufacturing, distribution, and consumption.

HACCP principles apply to risk analysis and monitoring procedures, corrective actions and verification, and record-keeping. Wisso’s embedded analytics application helps kitchens implement HACCP as well as comply with reporting requirements. For instance, it monitors the temperature of different devices used in food storage and preparation, and alerts kitchen managers to any problems.

Managers also use the application to track issues like food waste costs and related CO2 emissions, reasons for waste, and bottlenecks that contribute to waste. Real-time alerts also notify managers of pending waste so they can work proactively to prevent it.

However, the solution Wisso was using to embed analytics in the application was not cost-optimal for scaling analytics to many users.

Supporting scale

We wanted to give our client an embedded analytics solution that would cost less than their existing tool and allow them to scale with ease.

We evaluated three solutions and chose QuickSight for the following reasons:

  • Pay-as-you-go pricing would allow Wisso to cost-effectively scale to thousands of users without having to forecast user numbers and pay upfront licensing fees
  • QuickSight would work seamlessly in the AWS ecosystem the platform is hosted on, reducing implementation time and eliminating the need for physical infrastructure
  • QuickSight has an intuitive interface, allowing for a robust user experience and faster dashboard development with fewer resources

Solution overview

We migrated our datasets, assets, and content from our previous business intelligence (BI) tool to QuickSight in 5 months, including:

  • Frontend development for embedding and AWS Lambda development
  • Designing and creating six dashboards with new key performance indicators (KPIs) and drill-down support
  • Automating user authentication and row-level security so users can only see the data they’re allowed to access
  • Adding multilingual support for dashboard and reports in three languages, including reports, labels, and actual data

The data flows from IoT devices and end-user actions into Amazon Redshift. QuickSight ingests data from Amazon Redshift into SPICE (Super-fast, Parallel, In-memory Calculation Engine), which makes analytical queries process faster, reducing the number of queries and workload in data sources. It’s fast, plus there’s no additional cost on the number of times data is reused in SPICE.

Wisso’s embedded application uses Amazon Cognito for user authentication. QuickSight is embedded in the application, using dynamic embed URLs generated specifically for each user at runtime. Users are created automatically in QuickSight using Lambda at the first time they access the dashboard. It provides a seamless experience for users and requires no manual intervention. It uses row-level security to manage data access, so different users accessing the same dashboard will see the data restricted by the permission dataset rules. This means we only need to manage one dashboard for multiple customers. It saves us time and has accelerated our dashboard release cycle.

The following diagram illustrates the solution architecture.

Professional kitchen dashboard

The Professional Kitchen Dashboard enables restaurant managers to quickly monitor and respond to alarms in their operational controls across multiple locations. They can pinpoint which restaurants have alarms and delve into detailed data to identify the root causes. Staff can drill into detail reports from each dashboard widget for further insights.

The dashboards feature a comprehensive overview of various critical areas required by HACCP. They include tasks that display progress, identify late items, highlight trends, and present data in tabular format. Task alarms provide insights into both active and resolved alarms, showing which issues generated the most alarms along with their reasons and trends. Device alarms track appliance issues, such as high and low temperatures in freezers and refrigerators, while also detailing active and resolved alarms, trends, and reasons.

In addition, the cooling section monitors food cooling process trends to enhance shelf life. The hygiene segment evaluates hygiene levels against targets for items like plates, door handles, and knives, along with associated trends. Lastly, the waste analysis provides a thorough breakdown of total waste volume, categorized by customer, day, type, and ingredient, including the cost of waste and additional CO2 emissions, as well as the leading reasons for waste. All these insights are presented in clear tabular formats for easy reference.

Restaurant managers can use dashboards to monitor and forecast the food waste so that they can optimize the usage of ingredients in the processes of food preparation.

Managers also use dashboards and reports to provide documentation of food safety to health inspectors of issues that occurred and were remedied.

Results

In only 5 months, the number of dashboard users on the platform grew appreciably. Kitchen managers can use reporting to identify operations bottlenecks that contribute to food waste. They can also benchmark their results, for example, comparing their metrics to those of chain restaurants. Reports track key KPIs like waste, costs, emissions, and waste per plate or ingredient, as well as other critical functions like tasks, food hygiene, and device alarms. Critically, kitchen managers can also generate food safety reports for health inspectors, with deep traceability.

With Amazon QuickSight, annual spending on BI tools has meaningfully decreased. Additionally, it is faster and more robust than the previous system. It’s a great initiative that will help kitchens reduce food waste and their carbon footprint while creating safer kitchens.

Looking ahead

We’re looking at adding Amazon Q in QuickSight, a natural language processing option that lets our kitchen management teams get the information they need by asking questions. It’s an innovation that will make analytics more accessible for kitchen managers.

Conclusion

In this post, we shared how Wisso replaced an embedded reporting system in their client’s kitchen management application using QuickSight. With QuickSight, the client increased user adoption, saved on costs, and could implement new dashboards faster.

To explore QuickSight for embedding analytics in modern applications cost-effectively and quickly, see Amazon QuickSight.


About the Authors

Jussi Railakari is a Wisso solution architect for IoT, data, and BI. He leads the Amazon QuickSight analytics solution for Wisso’s kitchen operation management application. Jussi has over 20 years of experience implementing and maintaining of IT infrastructure, data platforms, and reporting solutions to on-premises and public cloud platforms. Jussi has worked the last 6 years on AWS IoT, infrastructure, and data platform technologies.

Anton Garvanko is a Senior Analytics Sales Specialist for Europe North at AWS. As a finance professional turned salesman, Anton spent 15 years in various finance leadership roles in supply chain and logistics as well as financial services industries. Anton joined Amazon over 5 years ago and has been with the Amazon QuickSight team for just over 2 years. He’s passionate about connecting the worlds of finance and IT by making sure business intelligence supports everyday decision-making across industries and use cases. Outside of work, Anton enjoys spending time with his wife and children, reading, and running.

Roy Yung is a Specialist Solutions Architect for Amazon QuickSight. Roy has over 10 years of experience implementing enterprise business intelligence solutions. Prior to working for AWS, Roy delivered BI and data platform solutions in the insurance, banking, aviation, and retail industries.