Hello HN,<p>I am entering a role which requires me to advocate for increasing the use of Data Science within business/industry for my state.<p>Data Science and Machine Learning became IT buzzwords a few years ago, and I've noticed that sometimes medium/large business will allocate funding for Data Science, but not really know <i>what</i> Data Science is, or <i>how</i> best it might add value to their business. Often the people making the spending decisions are business/MBA types rather than technical experts.<p>I am required to create and deliver presentations to CEO/CFO level executives on the value of Data Science. What are the best examples, the best 'wow' demos, the clearest case studies that show the value of these types of solutions? I want to get these people excited about Data Science / Machine Learning and the best way to do this is with concrete examples.<p>Please remember I'm not looking to train MBAs on how to perform Data Science - I'm looking to help these business focussed people understand how Data Science / ML might be useful within their organisation.<p>Thanks in advance.
Dataish person here.
If you search for Ml/ds you’ll get technical content that won’t help you much or fluff. You’d be better served by looking at companies you know do DS and searching for business problems they solved with it.
Examples: Amazon boosts sales by recommending products. Visa stops fraud by doing fraud detection. Walmart manages inventory by doing sales forecasting .<p>Once you’ve found a use case that could be relevant to your stakeholders you can dive into the technicals
I still think Kaggle's Titanic Survivability remains the best ML demo. Something everyone can relate to, surely. Find a challenge on Kaggle or AI Crowd that relates to your vertical and the business case will be obvious ;)<p><a href="https://www.kaggle.com/competitions" rel="nofollow">https://www.kaggle.com/competitions</a><p><a href="https://www.aicrowd.com/challenges" rel="nofollow">https://www.aicrowd.com/challenges</a><p>If you want to do a "live" demo, load custom data into Google Cloud AutoML, auto generate models and start querying. Everyone I've shown this to is amazed and immediately begins brainstorming!<p>Tackling High-Value Business Problems Using AutoML on Structured Data<p><a href="https://www.youtube.com/watch?v=MqO_L9nIOWM" rel="nofollow">https://www.youtube.com/watch?v=MqO_L9nIOWM</a>
At my company we put together a series of case studies on the use of Natural Language Processing within financial services.<p>Each case study started with explaining the business goal, then explained how they implemented it. Along the way, each case study introduces an NLP concept.<p>This e-book got quite a bit of traction. I'd say half the case studies would be relevant to other businesses as well, especially when you're looking at stuff like back office or customer services.<p><a href="https://www.fintext.io/how-finance-uses-nlp/" rel="nofollow">https://www.fintext.io/how-finance-uses-nlp/</a>
Read this blog:<p><a href="https://ai.googleblog.com/" rel="nofollow">https://ai.googleblog.com/</a><p>It maybe somewhat removed for what youre looking for, but heres the secret:<p>You cant speak intelligently and usefully about AI unless you have some kind of understanding about the cutting edge and whats possible. So read the newest research and connect it with your business
It's an intractable problem. Even sophisticated managers who are not experts will propose absolute crackpot ideas. Just show them some clips of Watson and AlphaGo then try to extract as much resources as you can.