- Data Talent Pulse
- Posts
- QnA with Jordan Potter: Avenues for Learning in Data Science, ML in eCommerce
QnA with Jordan Potter: Avenues for Learning in Data Science, ML in eCommerce
DTP #45: Gen AI and Data Readiness Challenges
This week:
QnA: We spoke with Jordan Potter, Data Scientist at SimplicityDX on the impact of ML on e-commerce and upskilling oneself in data science.
Data Readiness for AI: Only 6% of organizations have a production-ready application, highlighting data readiness challenges and the need for strategic changes.
AI in Nonprofits: 75% view generative AI as transformative for marketing, but lack of familiarity is a key adoption barrier for two-thirds.
💼 AI in Business
Gen AI and Data Readiness Challenges
An article from Harvard Business Review explores the factors that are impacting organizational readiness to implement Gen AI:
A survey reveals excitement about generative AI but lack of economic value realization.
Only 6% of organizations have a generative AI application in production, with many still at experimental stages.
Data readiness is crucial; poor data quality hampers genAI usefulness.
Data strategy changes are necessary, but most organizations haven't made them yet.
Companies are focusing on specific tasks like data integration, cleaning, and curation for genAI.
Prioritized areas for genAI development include customer operations, software engineering, marketing, sales, and R&D.
Despite excitement, 71% of CDOs prioritize other data initiatives over generative AI.
It's crucial to start preparing data now for genAI's transformative potential, despite competition for leadership within organizations.
💬 Q&A
The impact of ML in E- Commerce
We spoke with Jordan Potter, Data Scientist at SimplicityDX, on how machine learning is transforming the ecommerce landscape through chatbots and storefront optimization:
How would you say that machine learning is improving the e-commerce experience for businesses?
In a lot of ways. The big trendy thing right now is LLM’s, the foundation models, the ChatGPT's of the world. It is sort of an obvious case where you can have chat bots on your website. A lot of companies are using chat bots to do customer support.
People see that it's faster and provides a decent experience.
You can also use it in terms of building the site. One of the things that we're looking at is advertisements and building storefronts for advertisements. [Ex.] trying to see what an LLM can tell us about the advertisement that you've posted and how can we reflect that information into your storefront So, automation that may try to pull information from pieces that you know are successful and put it into the storefront itself.
The other thing where I think machine learning is helpful for ecommerce is reinforcement learning. Different reinforcement learning tasks to update a website in real time, whereas in traditional AB tests it would take multiple weeks.
Simple reinforcement learning can optimize many different types of websites in a lot less time.
On your second point, if someone clicks on an ad, they get through to the website and then ML can be utilized to make sure that their expectations from that ad are met?
One of the things we found is that a lot of advertisements just lead to like the standard product detail page on a website, like what you find on Amazon. A product of a shirt for example. The page may have the description of the shirt and other details, and that's great.
But an ad is a very different experience, where it's more about the feeling of the ad - the right imagery and colors and text. Maybe it's not even about the shirt itself but it is leading you towards the shirt. It may be more about the brand.
And so there's sort of a mismatch of experience going on. If we can pull in information from the ad, we might find that the person who clicked on this ad wasn't even interested in the shirt, maybe they're interested in something else. How can we incorporate that into the website that we build?
On a more individual note, what is your process when it comes to upskilling yourself? Do you focus on the domain that you're working in, or do you broaden your learning?
It’s a combination of both. For me, a lot of it has been looking into what tools I am going to need for a task and doing a lot of research to see what's out there.
So, for my current position, it is understanding the problems in terms of what's going on, and how can we then solve it? What attempts can we make to find a solution to this problem?
I tend to be more on the side of: Learn about a specific tool based on knowing that it can apply to what products we have versus you know, trying to figure out everything about AI or machine learning that's going on in the world.
I’ve found it more important to figure out the tools that I need for my job and then learning as a result of that.
🌐 From the Web
eBay's iOS app now includes AI-powered "Shop the Look" feature, allowing users to find and purchase items similar to those in uploaded images.
Google Workspace introduces AI Security add-on for $10 per user per month, enhancing protection with advanced threat detection, spam filtering, sensitive file classification, and post-quantum cryptography support.
Microsoft announced a $2.9 billion investment in Japan over two years for cloud and AI infrastructure, including skilling three million people in AI and establishing a research lab. Other tech giants like Amazon and Google are also expanding their data centers globally.
Generative AI's growth is met with cautious business adoption; ex. Reynolds American tests limited applications due to risks, echoing widespread hesitancy.
🏳️ AI for Good
More Than Half of Nonprofits Use AI according to Google Survey
Nonprofits worldwide face challenges due to limited resources and administrative burdens. Generative AI tools is poised to enhance productivity by 66%, offering potential relief for nonprofits to focus on impactful efforts:
Insights from a survey of 4,000 organizations in Google for Nonprofits program:
75% see generative AI as transformative for marketing efforts.
Lack of familiarity with generative AI is a major adoption barrier for two-thirds of nonprofits.
40% have no staff educated in AI.
Google for Nonprofits donated over $2 billion in Google Ad Grants and Workspace access, enabling 300,000+ nonprofits to access AI-powered technology.
Nonprofits seek simple, low-cost training tailored to the social impact sector for generative AI adoption.
A productivity guide accompanies Gemini for Google Workspace, offering AI tools for grant proposals, communications, image generation, donor databases, etc.
🤖 Prompt of the week
Act as a SQL code optimizer. The following code is slow. Help me speed it up: [Insert SQL] See you next week,
Mukundan
The Data Talent Pulse is brought to you by TeamEpic, a trusted global AI Talent provider. Learn more about us here.
Reply