Google Cloud delivers quicker deployment and scalability at Fluidly
UK-based Fluidly provides cash flow solutions for thousands of SMEs, via a platform that simplifies cash flow management through a unique combination of AI and financial modelling. Founded in 2016, the company quickly found that its original infrastructure had reached its limit and opted to migrate to Google Cloud Platform (GCP) due to its ease of use and AI functionalities.
Fluidly co-founder and Chief Data Officer Johnnie Ball says: “We feel the tooling around machine learning and AI is head and shoulders above the competition with GCP.”
Migration was soon underway, starting with Fluidly’s machine learning and data pipelines. First, data had to be extracted, aggregated and computed before being processed with Cloud Dataflow and pipelined to different databases. Cloud SQL was used to securely manage data, while BigQuery and Cloud Memorystore enabled the company to use the REDIS open-source protocol.
Training data for Fluidly’s AI models is in Cloud Storage, before being taken into the Cloud Machine Learning Engine, which forms the heart of its platform.
Regarding Cloud Machine Learning Engine’s role in the business, Johnnie Ball says: "The simplicity of the service and its flexibility makes Google Cloud Machine Learning Engine absolutely key for us. With just a command, we can train a model, add our secret sauce, and then serve predictions through an API. I don't think we could do that with any other platform."
Following the successful migration of AI and machine learning elements, Fluidly began to migrate the rest of its infrastructure, with the majority of services running on Google Kubernetes Engine, which offers ease of use and scalability.
Security is obviously a premium at a firm like Fluidly and the company built a bespoke encryption management system using Cloud Key Management Service. Fluidly also Google Data Studio for business intelligence reporting.
Since migrating, Fluidly has seen its deployment times fall from an hour to under five minutes thanks to Google Kubernetes Engine, while the ability to operate AI at scale enables it to service thousands of clients easily. The company is also able to train and deploy thousands of machine learning models, with a platform that streams 600GB of data daily.