Posted by

Elias Ayoub

on November 14, 2018

Share to

More Posts...

The Role of Analytics at an AI Company

by Elias Ayoub | November 14, 2018

AI is at the core of Rubikloud. It takes many skilled teams working in collaboration to apply AI to our users’ business problems in intuitive ways. We make products, but before any software engineers are involved in the process, we need to validate that our solutions fit our clients’ needs and are fine-tuned for their business and their data. The Analytics Team bridges the gap between AI and its business application, thus making us an integral part of the product development process.

Becoming subject matter experts of our data

As is the case with most modern organizations, the abundance and complexity of data presents unique challenges. As soon as new data enters the Rubikloud world, our team must interact with it. We take on the role of data ingestion for client pilots as these projects have a quick turnaround and the data we receive is smaller than what is required for a full implementation of one of Rubikloud’s products. Pilots are like demos of our product for new clients, and the Analytics team is involved heavily with these types of projects, but it takes a collaborative team across different departments to deliver a pilot. Our Client Solutions and Data Science teams help us to guide the company towards a successful pilot and the beginning of a long-term client relationship. When a new client is signed, we begin a full data ingestion process, marking the start of the product implementation process. Although it is not our team building the pipelines, we have, through years of experience working closely with our products, helped define the data requirements for our products. As data floats downstream, our team is validating the data and ensuring it arrives at its various destinations in order to make our products tick.

Making sure the AI is being intelligent

We feel that the fruits of AI are not being cultivated if our models are not being applied effectively. It takes a strong analytics team to arrive at and implement AI solutions. Our team’s close relationship with our clients allows us to deeply understand their high-level business objectives and their day-to-day operations. This type of relationship helps us guide our products towards new solutions. Our proximity to the Data Science, Product, and Client teams has become more defined over the years and now exists as a test-and-learn feedback loop. We report KPIs that we develop to these teams to help determine product and client success. The metrics we report provide insights to the Data Science team which supports their model improvement and ensures that Rubikloud continues delivering world-class products. Having a feedback loop like this is essential for driving innovation and improvement on the AI products we are creating.  

Analytics Feedback Loop

Ongoing Search for AI Applications

We have become experts in finding solutions for our clients that fulfil their business objectives by working closely to the intelligence that powers Rubikloud, and with the clients themselves. There are many ways to apply AI, but understanding which applications are feasible and which are not is crucial when building enterprise software. While our Product Engineering team works towards developing state-of-the-art features, the Analytics team leverages our unique position to prototype new features. Client pilots provide us with these opportunities since we are looking to build, package, and expeditiously deliver solutions that may be outside of immediate product capabilities. After delivery, we track our performance using relevant KPIs, and the successful measurements of these KPIs lead to an extended client relationship. At this point, our Product Engineering teams take over the work and develop product features from our beta solutions. This is how Analytics is constantly evaluating new feature opportunities and how those might address the common problems our clients face.

Fully understanding how customers are using our solution

By now, you should have a clear picture of the role the Rubikloud Analytics team plays in the end-to-end product implementation process. Ultimately, our role will transform into product analytics – focusing solely on how our clients are interacting with our products. This involves tracking usage patterns to better understand how our clients interact with the product and providing insights to our Product and Client Solutions teams. Since these insights help us understand which features are the most valuable to our clients and which ones are not used so often, it can help guide the direction of our products.

Rubikloud’s Analytics Team is integral to the product development process and, in many ways, helps guide the AI towards practical business applications. Product development is a marathon, and we all need to be going the same speed to stay on track with product roadmaps and client pilots. The KPIs developed by our team are important indicators of success, which guide our improvement as an AI product company, and are necessary for our growth.

We are constantly looking for sharp individuals to join our team and support the success of Rubikloud. If you think you would fit with our dynamic team, you should check out our careers page!