--- title: 'Project documentation template' --- Fraud as a Service (FaaS) === ## Solution Description FIS offers multiple products used by clients for detecting financial fraud. These products are deployed on FIS hosted servers and also in client datacenters which requires lot of effort to deploy, maintain and enhance these different products. Fraud as a Service project demonstrates fraud detection capabilities hosted as cloud based web service. It uses Azure REST API for web service and Azure machine learning service for fraud detection Due to time limitation, we focussed on Check Deposit Fraud, however, any type of fraud can be detected. Solution Features --- The FaaS web service allows clients to send transaction data using a web service hosted on Azure cloud and returns a score for the following check fraud indicators: * Seen Status * This indicator shows whether the deposited item account involved in the deposit has previously been encountered * Larger than Usual * This indicator shows how high the check deposit amount is compared to the median check amount observed for the account. * Exposure Risk * This indicator shows whether the deposit account is unusually exposed, i.e. the potential overall loss for the account associated with this deposit is unusually high. Technologies and architecture used --- Following technologies were used in this project: * C# * .NET core * Python * Azure REST APIs * Azure machine learning (ML) * In-memory database What is your code designed for --- The fraud as a service (FaaS) web service can be used by any FIS product that requires fraud detection capabilities. It can also be integrated with FIS CodeConnect to allow clients to invoke different fraud detection APIs directly. What is your code written in --- Fraud detection web service is written using C# and .NET core and is hosted on the Azure cloud. The fraud detection model is written using Python and Azure Machine Learning studio. Open source or proprietary software used --- Python Why it's cool --- Fraud as a service (FaaS) allows the capabilites offered by multiple FIS products into one product offering. That will result is huge cost savings compared to developing, deploying and maintaining multiple different products. Hosting the web service on cloud also offers capability to scale up performance based on demand by using cloud enabling technologies like dockers containers What's your Wow factor? --- Fraud as a service can be used to quickly deploy cross-channel fraud detection solutions using machine learning. Services can be added for different areas such as instant loan approvals, payment fraud detection, ACH/Wire fraud detection. We successfully demonstrated the ability to quickly learn and implement cloud based capabilities like Azure REST APIs and Azure Machine Learning during the time allocated for Innovate48.