The Boring Enterprise Nerdletter #10: Data Leak, Digital Transformation, GPT-3 API, Integration Architecture
Hi there,
We are excited to present to you the 10th issue of our Nerdletter! If you like to read our words, check out our voices on the all new Nerdcast, available wherever fine podcasts are sold.
As you will see below, we live in times of Great Uncertainty. But one thing you can be certain about is that in 2 weeks another issue of the Boring Enterprise Nerdletter will arrive in your inbox. And it will have exactly 6 non-dull stories. That’s one thing you can depend on!
-Jelena and Paul
Your Data Is Leaking!
As this issue was going to the digital press, news broke out about a potential data leak on the SAP Influence website. The site has since been taken down and the lucky “influencers” have received the emails informing them about the incident.
Just like in the similar hiccup with Microsoft PowerApps last year, the leak involved the OData APIs that have been empowered just a tad too much. In his whistleblowing blog post, SAP Mentor alum Tobias Hofmann shows how easily OData query can be manipulated to get additional information.
OData APIs / web services have been the best thing since sliced bread for SAP development and I hope that this incident is not used by some overzealous SAP Security teams as an opportunity to effectively shut down all innovation. But, as they say, “with great power comes great responsibility”. This incident should serve as a reminder to all the OData developers to be cognizant about data being exposed. Last year’s SAP Community blog Being Extra Careful With Exposed Fiori / OData Solutions by Jose Sequeira should be getting many extra views, I hope. JP
You, Too, Can Be Digitally Transformed!
5 or 6 years ago I started to hear the term "digital transformation" tossed around. Consulting firms used it as a catch-all prescription for any project they pitched. But it didn't seem like there was a solid definition anywhere. Heck, I worked for a consulting firm that occasionally used the term - and I could never understand what that freaking term meant. It reeked of so much BS.
The phrase eased off a little but has stayed alive as an industry term. I bring it up because a recent piece in the Harvard Business Review helped me cobble together a better understanding. The authors describe four pillars of digital transformation:
IT uplift: existing systems, brought forward out of obsolescence
Digitizing operations: business processes optimized by new computer tech
Digital marketing: putting computers to better use in marketing and e-comm
New ventures: the shiny future things that lead to growth
Things clicked when I read about these pillars, because my various client projects fell neatly into buckets. My own personal cloud of confusion lifted ever so slightly. Digital transformation - if combined with one or more of these pillars - can make some sense.
But that doesn't mean that talking about "digital transformation" is all of a sudden a sensible thing. If someone is selling you digital transformation, you better darn sure dive deeper into what they're selling - because that phrase by itself still doesn't carry enough descriptive weight. Agencies need to refine their digital transformation messaging if they are going to continue using it. PM
Great Big Mess of Great Uncertainty
According to an old joke, a pessimist is the one who says “things can’t possibly get any worse” and the optimist says “sure they can!” By that measure, the article Forget the Great Resignation, we’re in a Great Big Mess is very optimistic.
The authors describe the apocalyptic reality that meets us “just as we’re peering out of our caves and rubbing our bleary eyes at a world recovering from a two-year pandemic”. Raising prices, wage inflation, project delays, even more supply chain woes, you name it. This article by McKinsey is slightly less gloomy but concurs that great uncertainty lies ahead.
My grandma survived WW2 and she often told me “the most important thing is that there is no more war”. And living through the nuclear drills of 1980s and the USSR collapse of 1990s gave me a different perspective on the apocalyptic spectrum. If we can avoid nuclear war, we will pull through, I’m quite certain about that.
The silver lining of the Clouds of Doom could be the real push for sustainability, the rise of smaller but not less cool projects, accelerated adoption of all the AI, low-code, etc. tech that we’ve been writing about. The key is to follow the scouts’ motto and “be prepared”. JP
Sustainable Skynet
I've listened to quite a few of the Enterprise AI videos put out by Aaron Back at Acceleration Economy. Many of them are the short "Enterprise AI Minute" format, which is great because I like getting a bite-size trickle of thoughts that I can squeeze into other parts of my day.
Late last week Aaron offered some thoughts on using AI initiatives to help organizations reach their sustainability goals. He pointed out the potential of AI to add tons of cost-saving value to some of the large initiatives in renewable energy and battery storage. That could mean trillions of dollars for even more sustainable stuff. But he also points out that sustainable initiatives don't have to be trillion-dollar mountains. Business processes and physical workflows that use lots of power can be examined and optimized through AI even down to a single company facility.
I'm glad the smaller initiatives play a role in this discussion. It's not enough to push off all the sustainability work to large-scale projects. Businesses of all sizes have existential incentive to improve their own sustainability in-house, and new businesses to help that along will - hopefully - be an amazing growth market in years to come. PM
Reading: SAP Integration Architecture Guide
SAP recently published an updated document ‘Integration Architecture Guide for Cloud and Hybrid Landscapes’. At 100 pages, with many diagrams and tables, the guide is easy to consume not just by the architects but by anyone interested in the integration subject.
I was expecting a fluff piece on SAP Integration Suite and was pleasantly surprised that the document offers a good overview of various integration options. It covers API management, master data integration, “thing” and UX integration, as well as the products we wrote about in the previous issues: SAP Event Mesh and SAP Graph. I’ve also learned a new abbreviation: A2A for “app to app” integration, "chaining" of business processes and transactions of sorts.
It would’ve been super useful if the document was published in a PDF format. But even with SAP's quirky online reader, it is worth your while. JP
Stop Talking About NLP Already
If you've seen any past issues of the Nerdletter, you know we love AI that messes with words. At this point, any reasonable predictive analytics would tell you that you're likely to see another little story mentioning AI and natural language processing (NLP) in this issue. So let's just skip all the repeat-shaming and jump right in.
OpenAI released an updated version of the GPT-3 API that allows for editing and inserting AI-generated info into existing text content. Before this, most uses for GPT-3 were for adding text to the end of a prompt, based on what came earlier in the 'conversation'. Some of the examples were really mindblowing:
Take a short Python function and rewrite it in JavaScript
Take that JavaScript and turn it into a one-liner
Convert a simple list of information into JSON
Convert that JSON into YAML
Programming languages seem to get the most immediate benefit from this featureset, at least to start (users of CoPilot will already be familiar). To me, these features still represent tiny sparks of possibility in NLP that will surely, eventually grow to a roaring fire of true human communication enrichment. PM