Hey there!
Perhaps you have already heard about the data maturity of a company before. The term describes a measurement of how "data-driven" a company is already operating, from barely collecting any data in a structured way to making data analysis the foundation for basically every business decision.
Lately my mind has been busy (along with still studying for the MLOps Databricks course) considering the different stages of AI/ML maturity and what we as ML professionals can do to help increase this for our current teams through communication...
đź’ˇ The problem with AI on a tech-island
Many companies in the last 2 years have set AI goals, because it seems they want to hop on the bandwagon, the hurtling train, of AI progress. “AI is taking over the world, quick, don’t leave us behind!” type of energy. In their enthusiasm, they are like “we want to implement all this new technology”.
But the question is: Do you want to implement AI into the business or just as a project?
When you begin an AI project, it’s tempting to focus on the technical side. You hire machine learning experts, set up servers, and dive into coding. Your first project might be a chatbot, leveraging pre-trained models like GPT-4. It seems simple: no need for extensive training data or clearly defined tasks. But here’s the catch: without a clear business integration, your AI project remains just a tech experiment.
The illusion of progress
Once your chatbot is live, you might feel accomplished. Your company now has an AI project, and employees can use it. But is this really progress? In reality, what you’ve built is a web application, not a transformative business tool. The chatbot is just an app, using APIs and databases like any other software project.
The Solution: Integrating AI into Business - but how?
To truly use AI’s potential, you need to integrate it into your business processes and use it as a tool to reach existing business goals instead of making "use AI" another independent goal. This requires collaboration between tech teams and business leaders, basically you need to build bridges from your island to others (everyone working in corporate knows this is easier said than done).
Ideally you can find at least one department outside of IT that is willing to collaborate, then you can go through the stages of building a positive example:
- Data Collaboration: Work with the business teams to gather relevant data. For example, if you’re building a model to optimize ad spending, collaborate with the marketing team to access historical ad data and revenue figures.
- Cross-Functional Teams: Form teams with both tech and business expertise. This ensures that AI projects align with business goals and address real-world challenges. (This includes both teams being willing to educate the other one and learn from each other - again, often a huge challenge in fixed corporate structures and not to be underestimated)
- Pilot Projects: Start with small, manageable projects. Test AI solutions in real-world scenarios and measure their impact. Share successes and learnings across the organization so hopefully other departments will follow
If you have any personal experience with making this culture shift work, please hit reply with any tips - I would love to chat about this!
(I'm so serious about this, I really want to build more AI use cases but the hesitation is sooo real in so many situations and I want to know how to help)
🔍 Learning Resource of the Week
During my initial Google search on the topic, this blog post seemed the most honest and insightful, breaking down not only the different maturity stages (that are notoriously hard to pin down) but also the different dimensions. These dimensions - differentiating between tech maturity and ecosystem/organization/mindset is the crucial puzzle piece to the above problem of "AI as an island" in my opinion, but read for yourself:
​https://nexocode.com/blog/posts/the-ai-maturity-model-how-to-move-the-needle-of-digital-transformation-towards-an-ai-driven-company/​
​
Quote of stage 2: "Most organizations are at this level of AI maturity. They are just playing around with AI."
A reality that is hard to acknowledge in a business meeting but crucial for moving on to stage 3 and beyond
🎥 Project Updates
That vlog about studying for the MLOps course has been edited and posted to Youtube (marking my return from a 1.5 year absence from the platform...) you can watch it here:
​​
I also shared a bunch of videos in November about my day-to-day work life (including struggles on my business trip to Munich) on Instagram in a self-challenge to post more reels:
Catch you in the next update! Whether it’s next week or next year, I hope this inspires a little extra motivation to keep learning. As always, if you’ve got questions or just want to chat, hit reply!
Happy coding!
Sarah
