Anas Alfaris

Anas Alfaris, Co-founder & CEO of Intelmatix, talks to TBY about how the firm came to life, decision intelligence, and democratizing AI.

How did the idea of Intelmatix come to life?

Anas Alfaris: The idea for Intelmatix came about due to several critical gaps that we saw in the private and public sectors today. It is clear that enterprises of all sizes nowadays generate a lot of valuable data. A big trend in industry is the use of business intelligence (BI) tools, such as dashboards and charts that track internal KPIs and show a visual snapshot of certain trends. However, while BI dashboards contain a lot of information, they provide little in terms of intelligence. As a decision maker, the problem of indecision still remains. I would be dealing with information overload, without really closing the “last mile” gap between the data I see and the decision to be made. Figuring out how to manoeuvre through tons of information led to the concept of decision intelligence (DI). A discipline that became prevalent not too long ago, DI is intended to augment AI through the use of the social sciences and decision theory. Many businesses today want to utilize AI, but AI adoption still remains low. The reasons fall into four categories. First, the AI tools that exist today are highly complex by catering mostly toward the technical user, thus making them inaccessible for the non-technical decision maker. Second is the unaffordability of AI products. Most AI tools are still developed and targeted toward large enterprises; companies of small and medium sizes are usually left out of the game. Third, most AI tools are unscalable by design—they are a Frankenstein of libraries and software put together to solve a particular problem that often has a short timeframe. Fourth, most AI tools have limited usability; they are bound to certain functions of a business, without the ability to connect across the many verticals within the business. For these reasons, we thought of starting Intelmatix. We want to find a way to democratize AI to enterprises of all sizes, and offer AI technologies that are not only affordable, but also scalable, user friendly, and provide the decision maker with a clear recommendation at the end of the day. This is why Intelmatix was born.

What are your AI/DI offerings?

Anas Alfaris: We have designed Intelmatix in a way that it is able to serve organizations of all sizes and industries. We are sector agnostic, meaning that we serve companies from the banking and healthcare to the retail and manufacturing sectors as well. In terms of our offerings, we provide both solutions and products. Solutions are customized use cases that we develop end to end for a specific client, hence they are targeted toward larger enterprises. Products, on the other hand, are subscription based, generalized use cases that require light implementation, which makes them very affordable to small and medium enterprises as well. When it comes to products, we develop them specifically for repetitive problems faced by companies within the same industry. The idea is to tailor them to industries one by one.

Can you explain how your AI/DI products work?

Anas Alfaris: DI is at the heart of Intelmatix. Our products are designed around the decisions that the end user would be facing in the future while also building models that can account for future risks that may impact the business. We work closely with our partners in order to identify these use cases and we design our products so that the interconnectivity between these use cases are ensured. Imagine if, as a retail manager, I want to know when to order my inventory for the next week. To answer that, our DI platform would forecast sales for this week and the next. Depending on the current inventory levels and the lead time, it recommends the exact order time and inventory amount needed. Over time, the model can even estimate the exact effect of each marketing channel on sales. By learning the channel effect, our DI platform can not only recommend the optimal marketing strategy, but also feed this information back to the forecasting model and continuously improve the performance of inventory optimization. This will improve the next marketing campaign. Furthermore, by learning the demand in each branch at different times of the day and week, we can optimize the allocated staff for peak and quiet hours, accounting for any labor policy restrictions. Other common problems include figuring out where to open your next branch, or how that area will affect your sales over time. Our approach of paying close attention to the dynamics between the various use cases, we believe, differentiates us from other players in the market. We want to be positioned as the global leader in DI. We are not only about AI; DI is our bread and butter.