Placing customer data privacy at the center of their product, Moody Month's founders adopted a subscription-based model to avoid selling customers' sensitive health data for advertising revenue. They also made security a key consideration as they searched for a cloud provider on which to build their app. "The main decision was around how easily we can handle security and keep our customer data safe," Ferguson recalls. "With Google Cloud that's all built in."

Being able to store encrypted authentication details in Firebase Authentication while isolating sensitive user health data in Firestore meant Moody Month could keep user data secure. The fact that Google Cloud offered end-to-end data analytics solutions, including BigQuery and Looker Studio, also meant the company could keep its customers' data safe in a single environment, rather than exposing it to third-party software. This integrated data and analytics ecosystem would also enable Ferguson to work easily with the data, with robust APIs and integrations making it easy to connect different data sources, and near real-time processing enabling fast data access and insights. "The biggest unlock to innovation and great features is having data easily accessible to the people who need it," Ferguson explains. "Being able to achieve this in Google Cloud was a huge factor in our decision."

Moody Month also benefited from the Google for Startups program, which offered valuable business strategy support, as well as avenues to build brand profile. Google Cloud credits, meanwhile, allowed the team to build in a cost-effective way. "As a small startup, we have had to be super lean and efficient. Google Cloud credits have been completely transformative for us, giving us a way to grow our business sustainably," explains Thomson.

Thomson and Ferguson used Google Cloud to build Moody Month, their daily health and wellness tracker. Users check in daily to receive a personalized hormone forecast based on their cycle and insights from extensive research. Users can log their current symptoms, enabling the app to refine its insights over time, and the app offers research-backed advice tailored to the user's specific hormonal state and symptoms, to improve their well-being.

The core of Moody Month's predictions lies in its accurate cycle-prediction model, a microservice running on Cloud Run. This model, written in Python, uses machine learning to forecast a user's cycle length, crucial for delivering precise hormone forecasts. User-specific data, including symptoms and synced wearable data, is isolated in Firestore for enhanced privacy, while the machine-learning models themselves are securely held in Google Cloud Storage. For robust data analysis and ongoing model refinement, the outputs from Cloud Run services are streamed into BigQuery. The data is then visualized in Looker Studio, providing Moody Month's data team with a comprehensive and near real-time view of model performance and user insights, vital for continuous innovation and accurate forecasting.

Moody Month is built in a way that avoids exposing sensitive user health data to Large Language Models (LLMs), enabling it to deliver personalized health insights without compromising data privacy. 

To do this, Moody Month has developed an internal platform that uses LLMs to generate health insights from decades of health research and academic papers, which are verified by medical experts. Those LLMs, including Gemini 2.5 Pro, are then used to generate corresponding code that can detect these insights in user data, meaning the LLMs themselves never touch actual user data. When a user requests an insight, their private health data is processed by a separate Cloud Run service within Moody Month's custom platform. This platform then applies the pre-generated, expert-verified patterns to the user's data, performing sophisticated pattern recognition without exposing sensitive information to the LLMs. Once a pattern is matched, the corresponding insight is delivered to the user, allowing Moody Month to provide personalized insights without compromising user privacy.

Moody Month is now scaling its data insights platform, using Vertex AI to identify patterns across even more wearable devices and life stages. It also just launched comprehensive mental health support covering pregnancy, postpartum, and pregnancy loss, expanding its commitment to supporting women through every stage of their lives. 

For Ferguson, Google Cloud has been integral in getting the company where it is today. "We wouldn't have survived without the scalable infrastructure of Google Cloud," says Ferguson. "We have 10X-ed our user numbers over the past two years, using our original infrastructure. When we 10X again, it will all still work. There isn't any other platform within our business that is delivering that amount of value."