Transforming environmental services with data: How GFL Environmental uses HAQM QuickSight for operational excellence

This is a guest post coauthored by Vikram Singh, VP of Architecture and Development, and Aayush Patel, Director of Data Engineering at GFL Environmental Inc.

At GFL Environmental Inc., we’re proud to be the fourth-largest environmental services company globally and the only major provider in North America offering comprehensive solutions across solid waste management, liquid waste management, and infrastructure development. With operations spanning Canada and 20 US states, our over 9,000 employees serve approximately 4 million households and over 135,000 industrial, commercial, and institutional customers. In 2023, we achieved $5.57 billion in revenue, with forecasts projecting $6.2 billion for 2024.

In this post, we share how GFL Environmental transformed our business intelligence (BI) capabilities by migrating from our legacy BI tool to HAQM QuickSight, resulting in significant cost savings, improved performance, and standardized data governance across our organization.

The challenge: Scaling analytics for a growing environmental leader

As GFL continued to expand, we faced increasing challenges with our existing BI platform. We needed to determine the profitability of our various routes, identify underperforming segments, and take data-driven actions to improve productivity and efficiency. However, our previous solution presented several obstacles.

First, the cost was substantial, creating budget constraints while not delivering the scalability we required. Second, performance limitations prevented us from processing our growing data volumes, which had reached multiple terabytes. Third, our data processing was confined within the tool itself, limiting reusability across our AWS data lake components. Finally, we lacked proper governance around usage and best practices.

These challenges prompted us to seek a more robust, cost-effective solution that could scale with our business needs and integrate seamlessly with our AWS infrastructure.

Solution overview

After evaluating several alternatives, we chose QuickSight for four primary reasons: significant cost savings, superior performance with large datasets, the ability to standardize our dashboarding approach, and the forward-looking product strategy for QuickSight.

In collaboration with the AWS account team, we engaged in multiple architecture review sessions that covered AWS best practices. During these sessions, we conducted comprehensive technical deep dives into AWS Lake Formation, AWS Glue, and QuickSight, including its geospatial capabilities and security implementation featuring row-level security (RLS) and group-based access controls. The review sessions also covered HAQM Athena, providing a thorough understanding of the complete service ecosystem.

Our implementation began in July 2024, and within just 7 months, we successfully deployed over 15 QuickSight dashboards serving more than 1,000 users across GFL. Our data pipeline now processes approximately 21 TB of data stored in HAQM Simple Storage Service (HAQM S3), with AWS Glue handling data processing, Athena managing queries, and QuickSight delivering compelling visualizations.

The architecture integrates our GFL data lake with QuickSight through HAQM S3, AWS Glue, and Athena, providing a seamless flow of information. For security, we implemented single sign-on (SSO) through AWS IAM Identity Center, applying appropriate access controls across our organization.

The following diagram illustrates the solution architecture.

QuickSight offered a promising roadmap that stood out during our selection process, and we’re enthusiastic about working alongside AWS to influence the tool’s future direction.

Transformative dashboards driving business value

Our QuickSight implementation has delivered several critical dashboards that provide actionable insights across our operations:

• Route profitability – This dashboard analyzes route performance and identifies cost anomalies, helping us optimize operational efficiency, reduce costs, and enhance customer experience. It delivers insights on key metrics, including Revenue on Route, Yards, Disposal Quantity, Net Revenue, and Gross Operating Profit (GOP) per hour.

• Fuel consumption – This dashboard processes Geotab data to show fuel consumed and distance traveled by vehicles across North America. It enables our fleet management team to track performance, identify inefficiencies, and maintain compliance with sustainability goals.

• Revenue ranking – This dashboard helps identify services that can be optimized from a revenue perspective. Metrics like Price Per Yard help our sales team adjust pricing strategies effectively.

• Customer profitability – This dashboard delivers insights into the profitability of roll-off operations, helping optimize pricing and resource allocation. The map visualization shows profitability by zip code, city, and province, highlighting areas with low and high margins.

Realized benefits: Quantifiable improvements

The transition to QuickSight delivered substantial benefits across multiple dimensions:

• Cost optimization – Our transition to QuickSight has resulted in a 76% annual cost reduction compared to our previous BI solution—other BI tools had huge license costs associated with them. Additionally, per user cost was also high. This dramatic cost reduction has freed up resources for other strategic initiatives while providing superior capabilities.
• Performance improvements – QuickSight uses two data sources: SPICE (Super-fast, Parallel, In-memory Calculation Engine) and Athena datasets. The system currently processes 3 TB of data through SPICE, and the remaining datasets are accessed and queried directly through Athena. QuickSight has enabled us to seamlessly analyze over 3 TB of data with high performance for more than 1,000 users. Our previous solution struggled with 1 TB, whereas QuickSight efficiently handles our 21 TB workload. Currently, we have approximately 250 active users per day accessing 60 maintained datasets with 3 TB of SPICE usage refreshed twice in a day.
• Standardization and governance – With QuickSight, we’ve established standardized approaches to dashboards and data usage. Because our data lake resides on AWS, different components can be reused as needed, creating consistency across the organization. We’ve also implemented governance around usage and best practices, providing data integrity and reliability. Our access control system implements row-level security through user-based rules.
• Operational impact – QuickSight reports have significantly enhanced operational impact across multiple business units. Several BUs were able to monitor historical route metrics and identified opportunities for route optimization that increase efficiencies and reduce costs. Additionally, we saw business unit profit margins improve due to better visibility into key performance metrics, and Treasury is now able to proactively target consistently late-paying customers, helping to streamline collections and improve overall cash flow.
• Expanded user base – Starting with a small user base of 20 QuickSight authors and 70–80 readers, we’ve experienced significant growth to reach 1,200 total users today. This includes 30 authors and approximately 1,170 readers. We plan to further expand our user base to include thousands of readers by the end of 2025.

Custom visual embedding

QuickSight’s Custom Visual Embedding feature, combined with QuickSight Parameters, allowed us to seamlessly integrate a custom-built web application within our QuickSight dashboards. Using parameters, our operations team can dynamically pass key filters, like route IDs, date ranges, or vehicle IDs, from QuickSight to the embedded web app. The app renders interactive, layered visualizations, including vehicle GPS paths, stop durations, customer sites, and disposal facilities, directly within the dashboard.

The integration provides a unified experience, helping our operations team quickly identify inefficiencies like route deviations, prolonged stops, or unusual activities – without switching between platforms. The result is a streamlined workflow, allowing our users to analyze route audit and profitability data with enhanced spatial context, improving decision-making for fleet optimization.

Looking ahead: Future initiatives

Our journey with QuickSight is just beginning. We have several exciting initiatives planned.

First, we’re preparing to embed QuickSight dashboards in our InCab systems in Q1 2025, bringing analytics directly to our field operations. Second, we’re exploring HAQM Q in QuickSight to enable business owners from finance and operations teams to ask questions in natural language about route profitability, revenue forecasting, and resource allocation decisions.

We’re also expanding QuickSight for additional use cases across the organization and enhancing our geospatial visualization capabilities for route optimization. As we continue to grow, we expect to onboard approximately 2,400 additional users to the platform.

Conclusion

The implementation of QuickSight has transformed how GFL Environmental approaches BI. By providing real-time visibility into fleet performance, route inefficiencies, and service delays, QuickSight has become our single source of truth, improving data accuracy and consistency across departments.

Our standardized approach to dashboards and data usage has eliminated the siloed analytics of the past, and the cost savings and performance improvements have delivered immediate ROI. Most importantly, QuickSight has democratized data access across our organization, enabling teams at different levels to make informed decisions that drive operational excellence.

As we continue our growth trajectory, QuickSight will remain a cornerstone of our data strategy, helping us deliver exceptional environmental services while optimizing efficiency and profitability. The partnership with AWS has been instrumental in our success, and we look forward to continuing this collaboration as we evolve our analytics capabilities.

About the Authors

Vikram Singh is VP of Architecture and Development at GFL Environmental, having joined in 2022. Vikram is responsible for new applications, infrastructure, and modernizations on AWS, working closely with line-of-business executives to deliver technology that drives business outcomes such as route optimization, data-driven employee efficiency improvements, cost reduction, and customer service enhancements.

Aayush Patel is Director of Data Engineering at GFL. Aayush is responsible for building GFL’s enterprise data and location intelligence platform leading the data engineering and analytics team. With strong expertise in distributed computing and cloud infrastructure, Aayush plays a key role in advancing data-driven innovation at GFL.

Zaiba Jamadar is a Senior Solutions Architect working with enterprise customers in Central Canada. She joined AWS in 2020 and is passionate about helping customers through their digital transformation journey, as well as helping customers unlock meaningful insights using data and AI/ML. Zaiba enjoys public speaking and has presented at multiple AWS Summits and AWS re:Invent in the last 4 years. Outside of work, Zaiba enjoys traveling, music, and playing sports like badminton and tennis.

Mable-Ann Biewenga is a Principal Customer Success Manager working with global organizations based out of Canada. She joined AWS in 2021 and is passionate about accelerating digital and AI transformation for the organizations at scale. Mable specializes in people change management and driving business value realization through technology.