Over the last month, we’ve been working hard to enhance our machine learning capabilities and we’re excited to bring you up-to-speed on what’s been going on behind the scenes!

What’s new? – Ibex 1.2

Intelligent opportunity cost

Ibex is now even smarter on deciding whether to send a recommendation to a customer today or wait for a (potentially) bigger impact in the future. Previously this was determined by analysing each customer’s historic response to a marketing recommendation on each day of the week. With this month’s release we are incorporating opportunity cost calculations into the schedule models. This allows us to predict not just what the expected impact of a recommendation is today, but also what the likely impact would be in future – and therefore whether or not it’s worth being patient.

Unlimited scalability via Azure ML

Ibex is now using the Microsoft Azure ML Infrastructure. This highly scalable architecture gives us the flexibility to make predictions rapidly and further scale the number of simultaneous decisions Ibex can make.‍

Enhanced channel selection logic

We have enhanced the logic behind selecting marketing channels for customers and can analyse channel-specific costs. This can tip the balance between costlier channels(e.g. SMS) and cheaper ones (like email), while balancing these costs against opt-out rates and impact. 

‍