Machine Learning Series: Hear it from industry leaders (Pt. 3)
Part 3: Advice for business stakeholders featuring carsales.com Ltd and Culture Amp
This article would be helpful for executive leaders and decision makers wondering if ML would be a worthwhile investment for your business.
Previous articles
Part 1: Business applications featuring carsales.com Ltd.com Ltd and Culture Amp
Part 2: The journey of implementing ML featuring carsales.com Ltd and Culture Amp
Panel:
- Agustinus ‘Gus’ Nalwan, Head of AI and ML @ carsales.com Ltd
- Michael Ridgway, VP of R&D Engineering @ Culture Amp
Background:
carsales.com Ltd (ASX: CAR) is the largest online automotive, motorcycle and marine classifieds business in Australia. Gus is the pioneer of carsales.com Ltd’s award winning AI technology, Cyclops, an AI-assisted image recognition tool for vehicle recognition.
Culture Amp is the world’s leading employee feedback and developmentWorld’s top-ranked employee feedback and analytics platform. To date, they’ve worked with more than 2,000 companies and over 3 million employees. Michael was instrumental in making ML happen at Culture Amp, having built out their first ML team.
Q: What were some of your biggest learnings from your journey in commercialising AI at Carsales?
Gus @ carsales
“I’ve learnt that technical challenges isn’t the biggest barrier in deploying AI technology. The hardest thing is integrating it with the business and getting people to adopt it.”
Before we rolled out Cyclops to our mobile apps, it took me a while to onboard the other teams, understand their perspective, and integrate it with carsales’ other internal systems.
Before presenting an idea, the key is to thoroughly understand the product, and more importantly knowing the drivers and KPI’s of the stakeholders you’re pitching to. When your idea is aligned with their KPI, it will move much faster towards execution.
For any idea to be successful, I’ve learnt that you should involve everyone from the start, take them through the journey and let them be involved in helping shape the idea.
In rolling out something new, you also need to give everyone involved a clear plan of what will happen. This is especially important in mitigating any fear from the team. The people aspect again is the most important consideration in deploying AI because any misconceptions may result in a resistance.
Q: What advice would you give to executive stakeholders who aren’t sure if ML is a worthwhile investment for the business?
Michael @ Culture Amp
Data is the new commodity for this decade, and you’re going to be disrupted if you don’t explore what it can do for you. It depends on the type of business and where you’re at, but my take is if you have a lot of data and you’re only using it for one purpose, then there are many purposes yet to be discovered.
Our approach was to start small, so we can identify the good wins and assess how much resource to invest at each stage. We brought together a small team of data scientists and ML engineers, have them loosen up the data to see what they can come back with. I imagine as with any business, there are some capabilities we know we want and some we don’t know at all. It’s up to our data scientists and ML engineers to discover that.
Q: Do you see any difference in the way startups adopt ML to corporate businesses?
Gus @ carsales
The business models and priorities of a young startup compared to a corporate business are fundamentally different. Younger start-ups are still establishing a product-market fit and typically have a smaller customer base. The key areas where they could excel in are creative problem solving and leveraging cutting-edge technologies.
For a corporate business with an established customer base, it’s more about making processes more efficient and/or cutting costs. For those who are working in a corporate environment and are keen to explore ML, I highly recommend identifying smaller projects that can make a big impact, especially because building ML takes time. This approach is a great way to gain trust from the business to invest more in ML. ∎
Have some burning questions of your own that weren’t answered? Please comment below or let me know at rachel.chong@mitchellake.com as I’m planning the next series and would love new ideas!
