Ayasdi provides a revolutionary machine intelligence software that processes complex data, in overwhelming amounts, and helps to extract actionable insights from it.
It works with leading companies of various origins worldwide, including those from Fortune 500, and allows them to move from massive data to impactful solutions by means of topological data analysis, automation and machine learning.
Ayasdi has reached out to Perfectial to develop an application tailored specifically to the needs of financial data scientists and able to be integrated into the technical infrastructure of an enterprise. The idea was to launch a tool that, through Ayasdi’s predictive modelling, could help identify transactional frauds better, simplify trading decision making, and overall solve major problems businesses face.
Perfectial was commissioned to build, test and implement Modeler – an application that connects Ayasdi machine intelligence platform with end users and is easily installable on enterprises’ infrastructures.
Integrating the app with Ayasdi internal services and establishing a Single Sign On connection with their existing Workbench platform was especially challenging, due to Ayasdi’s unique infrastructure.
However, the result, the functional application turned out to be of immense value to financial analysts; it visualizes data, which has been processed by Ayasdi algorithms, and makes it much easier to comprehend.
A highly skilled cross-functional team of Perfectial engineers joined forces with Ayasdi’s in-house staff to fulfill following tasks:
- Develop from scratch the Modeler application design its simple yet efficient interface and establish continuous integration with Ayasdi platform
- Build Modeler’s pre-processing functionality so that the raw data entered by users could be turned into input appropriate for Ayasdi’s algorithms
- Make possible a custom, feature-based modeling to help data scientists build predictions based on criterias important to them specifically
- Include a semi-automatic modeling and workflow management features which allow users to:
- Manage and transform data
- Select Features
- Build Models
- Review Models
- Write scripts for automatic and prompt deployment of Modeler on clients infrastructure
- Include data post-processing into Modeler – users can have predictive models generated by Ayasdi compared them to the control data
- Analysis algorithms which help to understand deeply clients’ values and, consequently, provide more personalized offers
- Efficient, custom data modelling that allows firms to better assess financial risks, prevent frauds, etc.
- Built-in, detailed reporting that makes vast amounts of data easy to understand and interpret
- Access to machine intelligence software that helps build holistic customers’ profiles, based on subtle behavioral clues
There are a few things that really caught my attention. They (Perfectial) are very good communicators, which is absolutely a must, especially if you’re not in the same building, let alone in different time zones. If they’re facing an issue in a design area, they will have a very well-written JIRA ticket with concise information. It’s very natural and straightforward to understand what they want and to then respond.Ronaldo AmaEVP Product and Engineering, Ayasdi
Get in touch with us
Usually we get back in one business day.
Get in touch with us
We’ve received your message and will contact you shortly.