Page 29 - 10 Most Innovative AR VR Startups In 2019 up
P. 29
data warehouse initiative or don’t have the patience for the The role of Product Manager/
as part of an automation effort time it takes to deliver on AI/ Owner in Agile software
for customer management or ML projects. development is clear in software
marketing applications. The need development. Product Managers
for “labelled” (appropriately Given the uncertainty of success work as part of a team including
tagged) data in AI or Machine and timelines for many proposed UX, Engineering, QA and Project
Learning means that companies AI/ML projects, they may die Management. What seems to be
will also need to have a fairly before they even have a chance to missing in AI projects is this same
mature analytics and data capture begin. Add to this the newness of “Product” mindset. Good Product
infrastructure in place. While it the technology and the head-start Managers understand how to ask
may seem rational to approach AI dominance of Google, Amazon or find the highest value business
and Machine Learning as part of and Microsoft and the result is questions. Experienced Product
an overall IT data project, without that non-direct consumer-oriented Managers are experts at managing
an accompanying experimental/ enterprise companies may talk AI, the uncertainty of product delivery.
prototyping initiative, AI projects but start with Business Process Seasoned Analytics Product
can be buried as a consequence. Automation tools as their first Managers understand where the
Ironically, when the data is finally foray and wait for things to settle data is missing or buried. So, the
ready in the data pipelines, the (capturing data along the way for question of the day may not be
customer needs may have changed the future). “Where are all the data scientists
completely (ex: optimizing around we need?”, but rather “Where
CD distribution while the company Hypothesis #5: A lack of a are all the AI Product Managers
strategy moves to streaming). “Product” approach to AI/ who know how to ask the right
ML projects is core to project questions?”
Hypothesis #4: Companies failure and increased risk.
December 2019 29
