Crux Analytics
Partnering to build their first customer-facing web app and road test their proprietary algorithm
Project summary
Credit unions in the United States operate much like building societies in the UK: their members can borrow from pooled deposits at low interest rates. Many credit unions in recent years have made a push for more business members in pursuit of growth. However, they need outside help to achieve that. It’s risky to grow too fast and take on the wrong kind of members. So how can they tell the good businesses from the bad?
This is where our client, Crux Analytics, comes in. Crux provides hyper-local and highly specific analytics to banks and credit unions collated from a myriad of sources. This highlights areas of risk and opportunity for financial institutions to target their sales efforts and make informed lending decisions. The data Crux provides allows the credit unions to speak to small business owners with specific, intimate knowledge of their business context, improving the experience for both sides.
The brief
Crux is an early stage company with a proprietary algorithm developed by a super-smart founding team. This algorithm collects data on businesses and their owners through a range of APIs, analyses the risk against a business and its owners, and provides a rating back to financial institutions.
When Crux came to Loomery, this number-crunching was done manually and Crux wanted to test their algorithm at scale, in a customer beta programme, so they could automate important processes and needed a reliable partner to help them do this.
Building the MVP at pace
Loomery partnered with Crux to get their ‘minimum viable product’ (MVP) built and shipped at pace.
Collaborating closely with the team at Crux, we deployed robust infrastructure to host the web application, integrated with Crux’s number-crunching algorithm, and developed a secure front-end web application for their customers to interact with.
The beta participants can sign in to a self-serve portal hosted as a subdomain on Crux’s website and enter information about businesses they want Crux to provide data and insights for. This React-based web application then triggers Crux’s algorithm and compiles a custom PDF which can be downloaded from the web portal.
The solution
We leveraged the backend-as-a-service Firebase for a lean, secure hosting of the web application, database and cloud storage. This meant we could build the MVP quickly with a small development team of just two engineers. Firebase provides authentication out the box, shortcutting the work involved in user identity and login. We utilised Twilio SendGrid to send emails notifying customers and the Crux team that reports were ready for download.
There were challenges to overcome throughout this process. As an example, some of the APIs that Crux’s algorithm relies upon enforce ‘rate limiting’, which is throttling of high traffic from the same source in a short period of time. This could have been a real issue as Crux’s customer base grew, translating to lengthy delays for customers waiting to receive their results, or even failing requests. We introduced bespoke controls to circumvent these rate-limiting issues, staggering traffic to manageable levels.
The results
Loomery delivered Crux Analytics’ MVP web app in just two weeks. The MVP is live and being beta tested with customers. Over the coming weeks Crux will gain valuable feedback to help improve the algorithm and plan their next phase of growth.
"Working with Loomery, we’ve been able to get our MVP live and in customers’ hands quickly and efficiently in a very low friction manner, and without compromising on our needs. We’re able to focus our limited resources where they’re needed most - engaging with our customers, learning from having our product out in the wild and updating/improving our core technology as we go."
— Nathan Bennett, Co Founder
Related case studies
