From 1 Day to 4 Hours: Delivering Faster Insights on Emergencies for RapidSOS

Executive Summary

RapidSOS, a leader in safety solutions, engaged Mutt Data to modernize its data infrastructure, preparing for AI-driven innovation while immediately improving data freshness, quality, and operational efficiency.

Through a strategic migration to Databricks and optimization of ingestion pipelines, we reduced data lag from 1 day to 4 hours, and achieved significant cost savings, improving the responsiveness of emergency services. We also established stronger data validation, monitoring, and lineage tracking procedures to enhance analytics reliability, governance, and data quality.

About the Company

RapidSOS is a safety platform that securely connects life-saving data from over 540 million connected devices, apps, and sensors directly to 911 and first responders worldwide. Founded in 2012, the company aims to modernize emergency response systems by delivering real-time, incident-specific data that enhances situational awareness and shortens response times globally.

Introduction

In 2024, RapidSOS initiated a major transformation to enhance its data infrastructure and improve emergency response efficiency. Recognizing the need for a robust, real-time data system capable of integrating data from over 540 million connected devices, RapidSOS partnered with Mutt Data, combining hands-on execution with strategic data engineering expertise. These enhancements were pivotal in reinforcing RapidSOS’s leadership in emergency response services.

The Challenge

RapidSOS faced critical challenges in maintaining data quality, reliability, and timeliness. Their existing ingestion system was fragmented, with over 20 pipelines often relying on complex workflows, increasing operational risk and latency.

Data freshness varied widely, with some critical datasets taking between 24 and 72 hours to update, obstructing the ability to make timely decisions during emergencies.

Legacy orchestrators added operational complexity, causing occasional ingestion failures and making it difficult to scale for new data sources and geographies.

The Solution

Mutt Data led a comprehensive modernization of RapidSOS’s data infrastructure, implementing a robust, scalable system to improve data freshness, quality, governance, and cost-efficiency, while laying the foundation for future AI applications. The new architecture supports both real-time and batch data flows with greater reliability and performance.

In parallel, Mutt Data focused on two key tracks: migrating and stabilizing legacy workflows while developing new ingestion pipelines for incoming providers. We ingested raw 911 call data from various Emergency Call Centers (ECCs) across multiple U.S. states and counties, sourced from AWS storage, cleaned files, added new metadata fields, and enriched data in the CDR Calls table.

The migration of key ETLs to Databricks Workflows was a pivotal step in this process, significantly reducing failures in data ingestion, improving observability, and cutting data freshness lag from 24 hours to under 4 hours. By leveraging Databricks’ Job Orchestration, Delta Lake transactions, schema enforcement, and Auto Loader for scalable, incremental ingestion, we improved the reliability and visibility of RapidSOS’s data operations.

Across all pipelines, we implemented enhancements to boost quality and reliability:

• Near real-time data freshness validations.
• Data quality checks at ingestion and transformation stages.
• Comprehensive monitoring dashboards and alert notifications to detect anomalies, incorrect values, and inconsistent data early.

Beyond pipeline restructuring, we optimized data storage and processing by refining partitioning strategies and eliminating unnecessary transformations, enabling faster and more efficient data retrieval.

We also strengthened data lineage and governance, ensuring full traceability of data transformations while reinforcing quality checks at every stage of the pipeline.

Additionally, by introducing resource tagging strategies, enforcing cluster policies, promoting serverless architectures for short jobs, and right-sizing compute resources for memory-intensive tasks, we achieved a 5–15% projected annual cost reduction.

These enhancements not only improved the speed and accuracy of emergency response intelligence but also fostered a more integrated and collaborative data engineering environment within RapidSOS.

We modernized their infrastructure, strengthened operational workflows, and laid the foundation for AI-driven analytics, empowering RapidSOS to deliver real-time insights to first responders and scale its product capabilities.

Our close collaboration with the RapidSOS team was key to the project’s success. By adapting to their workflow and embedding ourselves into their routines, we became more than external advisors—we worked side-by-side toward a shared goal.

Impact

The improvements achieved through our collaboration had both immediate technical and business benefits: Reduced data freshness lag: The calls table update time decreased from 1 day to under 4 hours, providing RapidSOS with fresher and more reliable emergency data.

• Stronger operational workflows: Migration to Databricks Workflows enhanced orchestration resilience, reduced failures, and improved visibility across the ingestion lifecycle.
• Improved data governance and quality: We implemented enhanced data validation, monitoring, and lineage tracking processes, enabling more robust governance and greater trust in analytics outputs.
• Cost optimization: Serverless compute adoption and cluster right-sizing initiatives reduced processing costs by an estimated 5–15%.
• Data foundation for future AI initiatives: With better quality, faster refresh rates, and broader data coverage, RapidSOS is now positioned to scale the development of AI agents and predictive analytics.

These improvements not only enhanced operational efficiency but also strengthened RapidSOS’s readiness for scaling AI-driven analytics and predictive insights, ensuring the company remains at the forefront of intelligent safety innovation.

Wrap-Up

By modernizing RapidSOS’s data infrastructure, optimizing workloads, and setting the foundation for AI-driven analytics, Mutt Data fundamentally improved the accuracy, speed, and reliability of emergency response. These enhancements not only bolster RapidSOS’s leadership in emergency response technologies but also contribute to saving lives by ensuring that first responders have access to the most up-to-date information when it matters most.