10x improvement in time-to-first-token for AI responses
1.8-second average response time for natural dialogue
98.4% No-Harm Clinical Safety from patient interactions
Care Delivery Success increased from 73.0% to 82.3%
Quadrivia uses Gemini, alongside other LLM models, to automate up to 30% of clinical tasks, delivering 10x faster response times for patients.
Quadrivia uses Gemini, alongside other LLM models, to automate up to 30% of clinical tasks, delivering 10x faster response times for patients.
Finding the right AI provider to meet
growing healthcare demand
Global healthcare faces a pressing challenge: as patient demand grows, the number of available clinicians remains finite. This disparity means medical professionals often have less time to dedicate to direct patient care.
Dr. Ali Parsa founded Quadrivia to solve this challenge. Quadrivia’s clinical AI platform, Q, automates routine tasks, such as post-operative follow-ups, to transform how care is delivered. By managing entire patient journeys autonomously, from scheduling to medication management, Quadrivia enables clinicians to focus on complex patient needs and support a more productive patient experience, without increasing headcount.
Building an AI assistant capable of safe, real-time medical conversations requires a fast, reliable technical foundation. Quadrivia originally worked with a third-party cloud and AI provider, but soon needed a technology stack where every component was optimized for performance and security. The company chose to move to Google Cloud for its specialized AI infrastructure and joined the Google for Startups Cloud Program to help build and scale the platform.
“In this field, you’re as dependent on the security of the data and model performance as you are on your own team,” explains Dr. Parsa, Founder and CEO of Quadrivia. “We chose Google Cloud because we admired the pedigree of research from Google Health, and because its technology and customer-focused culture made it right for the long term.”
In this field, you’re as dependent on the security of the data and model performance as you are on your own team. We chose Google Cloud because we admired the pedigree of research from Google Health, and because its technology and customer-focused culture made it right for the long term.
Dr. Ali Parsa
Founder and CEO, Quadrivia
Engineering human-level performance with Gemini
With Quadrivia, clinicians can use the Q Manager to design and test medical journeys. For the clinicians building these journeys, speed is essential. If the system is slow to process a request, the design process becomes cumbersome.
To solve this, Quadrivia moved the reasoning workloads for the Q Assistant to Gemini on Gemini Enterprise Agent Platform. This resulted in a 10x improvement in time-to-first-token—the speed at which the AI begins to respond—compared to previous models. This technical leap changed the clinician experience. Instead of waiting for the system to process information, they can now talk to the platform in real time to adapt workflows or simulate patient interactions.
Because Gemini is multimodal, meaning it can process different types of information like text and voice simultaneously, clinicians can move between chat and voice messages in one session without losing context. While the assistant speaks an answer, the same content appears as text on the screen, allowing clinicians to quickly copy values or check names.
Privacy is a top priority when handling patient data. Quadrivia uses Sensitive Data Protection and Security Command Center to identify vulnerabilities and protect information. To ensure clinical accuracy, the platform uses Cloud Healthcare API to integrate with electronic health records. This provides the specific context needed to make informed decisions, such as identifying a safety signal based on a medical history. Clinicians can also use natural language to tell Q to route a call to a human if a patient mentions a specific symptom, making the platform easy to manage without technical training.
“We chose Gemini Flash because of it being a unique model when it comes to combining low latency and high accuracy. It is a fantastic model to carry out complex tool calling, deal with multi-modality and empathetically give answer to challenging situations with patients - without having to trade off speed,” explains Enrique Herreros Jiménez, Head of AI at Quadrivia.
We chose Gemini Flash because of it being a unique model when it comes to combining low latency and high accuracy. It is a fantastic model to carry out complex tool calling, deal with multi-modality and empathetically give answers to challenging situations with patients - without having to trade off speed.
Enrique Herreros Jiménez
Head of AI, Quadrivia
Automating the future of global healthcare
Quadrivia recently completed a large-scale trial using real-life cases. The results showed that, with the help of Google Cloud, Q can perform on a par with human clinicians for a wide range of designated routine tasks, successfully achieving a Care Delivery Success increase from 73.0% to 82.3%.
With the low latency of Gemini, Q delivers a 1.8-second response time, supporting human conversations by keeping interactions fluid and natural. As a result, Q has received a 98.4% No-Harm Clinical Safety from users. Furthermore, the data integration provided by Cloud Healthcare API ensures that Q has the right context at the right time, leading to a 99.96% clinical accuracy rate.
As Quadrivia grows, Google Cloud provides the scale needed to manage global operations. Using Google Kubernetes Engine (GKE), the company can deploy its services to reach patients in different geographies, including those in rural areas who rely on landlines.
The next step for the platform is to use the advanced reasoning of Gemini to create self-healing AI. This system will autonomously monitor conversations to identify and fix technical issues or delays, allowing Q to improve performance without manual intervention.
By using Google Cloud, Quadrivia has evolved from a startup with a bold idea into a service capable of reducing the clinical burden for providers worldwide.
“Healthcare has long been defined by scarcity, but with AI we can deliver an abundance of clinical expertise,” says Dr. Parsa. “With Google Cloud, we are combining a secure infrastructure with world-class AI to create a reliable clinical assistant that makes better care possible for everyone.”
Healthcare has long been defined by scarcity, but with AI we can deliver an abundance of clinical expertise. With Google Cloud, we are combining a secure infrastructure with world-class AI to create a reliable clinical assistant that makes better care possible for everyone.
Dr. Ali Parsa
Founder and CEO, Quadrivia
Quadrivia is a healthtech company that pairs customizable clinical AI with expert clinicians to handle repetitive routine clinical processes, end-to-end.
Industries: Healthcare and Life Sciences, Startup
Location: UK
Products: Cloud Healthcare API, Gemini, Gemini Enterprise Agent Platform, Google Kubernetes Engine (GKE), Security Command Center, Sensitive Data Protection
