...

Co-Innovation Challenge

RapidSOS is an innovator in emergency services technology, leveraging the capabilities of everyday devices to deliver critical, life-saving data during emergencies. By partnering with companies in home and commercial security, health care, automotive, telecom, and insurance, they provide information directly to 911 and other emergency response systems, increasing the speed and effectiveness of emergency services far beyond traditional 911 phone calls.

In an effort to further close the safety gap, RapidSOS joined forces with the Microsoft AI Co-Innovation Lab to develop a copilot for 911 dispatchers that could transcribe information from an emergency call and deliver relevant information back to the dispatcher using RapidSOS’s vast sources of partner data. They also wanted this AI assistant to generate conversation prompts and questions so that dispatchers could better assist people based on the nature of each call.

Working with numerous vendors, various data sources, and both structured and unstructured data, RapidSOS was challenged to determine how to design and build their solution. Together with our Lab engineers, RapidSOS’s team was able to bring this critical solution to life quickly while addressing its complexities for the greatest impact.

In the Lab

The Microsoft AI Co-Innovation Labs take companies’ complex AI challenges and collaborate with them to turn business use cases into scalable, working prototypes that those companies can bring to market. In the case of this solution, the Lab team huddled together to determine the best approach, communicating with RapidSOS to solve the technical challenges unique to their solution.

Coming out of an initial design sprint, RapidSOS and Redmond’s AI Co-Innovation Lab team built a solution architecture powered by Azure AI services that could ensure accurate and timely data gathering from multiple sources—ranging from home security to weather and traffic data.

RapidSOS 911 Dispatch Copilot solution architecture
RapidSOS 911 Dispatch Copilot solution architecture

In the rapid prototyping sprint, the engineering teams ramped up their efforts over four days, creating a “911 Dispatch Copilot” prototype that included transcription capabilities and generative questions and guidance.

The Lab team created a custom, hybrid approach to transcription with Azure AI Speech and Azure OpenAI Whisper, since neither technology could solve RapidSOS’s challenge alone at the time of the engagement. They also collaborated to create a gpt-4o model-based solution that would gather all available prompting information together, including speech-to-text transcription from 911 calls, data from RapidSOS’s partners, and RapidSOS’s own documentation. The model would then propose questions and offer guidance to the emergency dispatcher.

RapidSOS completed its engagement with a working prototype that was cost-effective and time efficient while retaining accuracy and performance. It could interpret highly complex and dynamic data and communicate relevant, life-saving information.

RapidSOS team at Microsoft’s Redmond office
RapidSOS team at Microsoft’s Redmond office

Solution Impact

RapidSOS’s 911 Dispatch Copilot prototype has the power to assist emergency dispatchers with rapid, accurate information and guidance, reducing the workload on emergency workers during high-stress and time-sensitive situations. This transformative solution can empower RapidSOS’s mission to close the safety gap using state-of-the-art Azure AI technologies and the expertise of the AI Co-Innovation Lab team.

Join Microsoft AI

Co-Innovation Labs today!

Apply now to get the opportunity to co-engineer your solutions with Microsoft Technology Experts.