As part of our aspiration to make advanced solutions that help tackle challenging business problems. We at believe that research is key in any Artificial Intelligence infused solution and therefore we work and collaborate with research centers to build state-of-the-art solutions.

Natural Language Processing | Transformation

Almost 90% of the data volume is in an ambiguous unstructured textual format that is hardly useful, and the issue faced is that this is the form most valuable insights take. Thoughts, opinions, ideas, and facts, although significant, can never be captured in their raw form by machines. Our first research area is a transformation stage where we aim to convert unstructured data points from organizations or open data into practical representative meaningful features that can aid decision-making.

Time Series | Temporality

Every data point that is captured or transformed by machines has a certain type of temporal dependence, when time is not factored in, data is considered to be disconnected and misplaced. Our second research area connects the dots and builds a relationship between data and time, hence the name Time-Series. We focus on embedding use-case data across time and the capability of building temporal relationships between the features and their target.

Multi-Agent Deep Reinforcement Learning | Decision-Making

When having a massive amount of valuable insights produced by machine learning models and infused by their temporal dependence, manual analytics or hard-coded rule-based systems fail. The issue peaks when the data is not relevant to only one entity but countless entities living in similar environments or competing in the same environment. Our third research area aims to automatically build rules through simulations and maximize predefined benefits so that we can provide evidential recommendations that aim to power our Sales, Marketing, and Customer-Success platforms.