TechBytes with Nima Mirbakhsh, CTO at Morphio
Tell us about the team and technology you handle at Morphio.
We have a nimble and highly motivated team of Digital Marketers, Software Engineers, and Data Scientists. We’re working together to change the way Agencies and Marketing teams approach digital data and decision making. Core AI technologies include TensorFlow for anomaly detection and forecasting. We also integrate a handful of ‘off the shelf’ scalable technologies and custom technologies that we have created to protect digital marketers against their unexpected vulnerabilities such as human errors, technology failures, and disruptive competition.
Data Science is incomplete without discussions on Big Data and Analytics. What kind of data and analytics do you operate with at Morphio?
Our AI integrates and operates with all Google, Facebook, Microsoft and LinkedIn Marketing analytic products. Morphio has created a new category of advanced analytics and insights including anomaly detection. We employ Machine Learning and a predictive model that includes normalized historical data that can be viewed as daily, weekly and monthly trends – plus a few other parameters to help digital marketers make strategic budgeting and tactical optimizations much faster than previously possible.
In the last 5 years or so, what kind of unique Deep Learning capabilities emerged? What kind of products truly showcase your technical maturity at Morphio?
Roughly 10 years ago, many people in academia were skeptical about the efficacy of Neural Networks. Today Neural Networks are becoming much more common and are significantly changing the way we live such as the technology behind Google translation, Amazon Alexa, and iPhone face detection.
An interesting example of our technical maturity is how we approach target audience segmentation. Originally we were using clustering algorithms to segment and profile the best and worst converting audience segments, but we figured out a more fruitful approach. While I am not prepared to expose our secret sauce at this time, I can assure you that clustering was not the best approach.
Tell us more about your work in Explainable AI and the kind of adoption you are witnessing within your customer base?
In general, our customers are now much more knowledgeable and realistic about AI compared to a couple of years ago. Both Google and Facebook have been pushing digital marketers to embrace AI and Machine Learning which has been helpful for the overall awareness of AI, but they are both very ‘black box’ as they compete to own a bigger share of ad spend.
Morphio, on the other hand, is very interested in educating our customers about our approach to AI. We feel that there is a necessary balance between humans and machines. While we do not want to give too much away to our competitors, we feel that our customers do need to understand how we are thinking and arriving at data insights so they can re-think their team structures and create new workflows and processes to scale more smoothly and efficiently.
Currently, our approach to Explainable AI is typically done in real-time with our customers when they invite us to look at their data where we will talk, investigate and plan as a group.
What are your predictions on the future of Digital Marketing, especially with enriched data and analytics delivered to the Marketing teams in real-time?
I think the most common fear about AI is that people (primarily Hollywood) paint a picture of AI and technology replacing humans, but in fact, AI is being adopted as a supportive technology – given the global unemployment rates. My prediction about the future of Digital Marketing is more of a level playing field between small and large teams as access to advanced analytics and insights is automated. Historically data-driven thinkers and leaders have had an edge, but as advanced analytics and AI such as Morphio are made available and adopted I think we will see more focus and success coming from Marketing strategy along with genuinely new creative approaches.
Tag the one person in the industry whose answers to these questions you would love to read
Nima Mirbakhsh leads the Morphio Product and is a Computer Scientist who specializes in research on Data Mining and Machine Learning.
Nima received his PhD in Computer Science from Western University in 2015 and has published multiple referred papers in international journals and conferences since 2009.