The Boring Enterprise Nerdletter #12: SAP Innovation, ABAP, IT Skills, Ansible, DALL-E 2
Hi there,
In a totally uncoordinated effort, we went "all in" on the enterprise memes in this issue. There is probably not much time left before AI starts generating these memes and then hacks your laptop camera to watch your facial expression and gauge the amusement level.
But today is not that day. Please enjoy this Nerdletter issue, proudly made by humans, for humans!
-Jelena and Paul
SAP Innovation Awards SPOILER: No Slaps
Every year SAP sponsors an innovation contest among its customers and partners. As you might expect, it centers around SAP solutions in use productively. I'm curious how many entries they receive, since this year there are 60 finalists and 25 winners.
Two of the winners stand out to me. First, this great edge computing/AI mashup from Avangrid. It uses cameras on the utility's vehicles to gather images of poles and wires, then edge computing devices inside the trucks use AI to suggest images that contain probable defects in the equipment. This saves inspection time, and further will produce tons of image data that can be analyzed later for more use cases. It's in that sweet spot of relatively new tech being applied directly to a business problem. So cool.
The other standout is energy distribution company Netze BW using Apple watches to give workers quick, safe access to SAP work orders. Workers can be more efficient with up-to-date data, while seeing safety benefits through automatic triggers for detected falls. But the reason this particular use case really stands out to me is that I had the chance to work on something very similar with Android Wear about 7 years ago.
Don't misinterpret my "I did this way back when" as pouty negativity! Finding and applying human value to tech always takes time. In my earlier tech-oriented years, I had to change my perspective on business software: don't come to any enterprise vendor's innovation awards looking for tech that's just peeking out of MIT Labs. Come looking for stories, and get inspired to write your own. PM
ABAP Developers Are Developers Too
This Nerdletter issue will hit the digital newsstands shortly before AIM 2022 event, where I’ve been invited to speak. The subject of our discussion will be a recent SAPinsider survey on the SAP application strategy and development. The survey indicated that there is an abundance of ABAP on-premise skills and at the same time, severe shortage of ABAP in Cloud, Cloud apps on SAP BTP, and other buzzwordy app development skills.
Is there though? To borrow from another meme, “they are the same picture”: ABAP developers are developers too and they are perfectly capable of using other programming languages and platforms. (And ABAP in Cloud, for example, is not even that different.) Of course, there will always be “another Excel file upload program, another dollar” crowd, but let’s not use that as a benchmark. The fact that ABAP developers have not yet become web and Cloud savvy en masse has more to do with the “inadequate clarity of business needs and priorities” of SAP customers (another survey finding) than with the developers’ lack of will or intelligence.
There are many bright, passionate, and skillful ABAP developers who want to do more than just RICEF. How do I know? Because I’m interviewing them for their new jobs. They are leaving the companies that don’t support their growth and [checks notes] suffer from “inadequate clarity”. Developers want to upskill but we also want to make sure it is meaningful and creates value. Perhaps what SAP customers and partners need to solve a “skill shortage” problem is to take a long hard look in the mirror. JP
Only Grandpa Knows How To Run The ERP
A piece from Nick Denning in Enterprise Times asks if we're facing an IT skills shortage in business tech. Put succinctly: "IT professionals with the skills needed to keep businesses running are disappearing from the world of work". He also calls out a Computer Weekly piece, which declares that CIOs need "people who can manage ERP or CRM or HCM projects. Who can convey the importance of technology and explain tech jargon to business people."
Schools and boot camps have been focusing on the sexier, digital frontend skills like "apps, mobile, AI, and data mining" - but Denning wonders if that starves the business systems of what they need? His proposed way forward is one (or more) of "maintain, modernise, migrate". To maintain, make work more attractive to those retiring early - like benefits that emphasize work-life balance and skill-transfer programs for younger folks. Modernize systems by refactoring business logic into API services that can work well with the latest GUI tools. Finally, migrating systems means - you guessed it - taking components of systems to the cloud whenever possible.
I'm of two minds when it comes to oldsters and skill gaps. (Now that I'm 40 I'm fast joining the oldster ranks.) On the one hand I can see that fewer and fewer young IT folks are climbing into the enterprise software space, especially development in SAP - at least in the US. On the other hand, when I get the chance to work with those few younger folks who do leap into enterprise, I am blown away by their capabilities and willingness to challenge the status quo. Young Paul was never that awesome. PM
Think THEN Automate
Red Hat’s webinar The value automation brings to SAP took place on March 10th but all the sessions are currently available on demand. Unsurprisingly, it was focused on selling Red Hat’s Ansible Automation Platform.
Despite generic “automation platform” name, Ansible is more about “moving your SAP workloads to public clouds” and server scripts than about automation of sales order creation from an email, for example. So, if you are wondering if Ansible is something like SAP iRPA then the answer is "not really".
One thing I liked from this event though was a passionate presentation “No successful SAP modernization journey starts without automation” by Belkacem Moussouni. The title reminded me of the Volkswagen’s famous “Sign THEN drive” commercials because automation really should be preceded by some strategy and thought process. Otherwise, it will just end up doing stupid things very fast. I also chuckled at the mention of “Automation Architect” that will soon exist in every company. We hear the same thing in other presentations on every subject. With the Data Architect, UX Architect, Integration Architect, and all the other Architects, soon SAP customers will need bigger offices. But jokes aside, the presentation correctly took a jab at “too many unintegrated, domain-specific tools” and noted that automation is not as much about pure cost reduction as it is about “focus on high value-add work”.
TLDR: If you’re looking for a passionate plea for automation, check out this session. And if you are actually interested in servers and workloads, then go see How Red Hat Ansible Really Works demo. JP
Are You In The Mood For Emotional AI?
When I saw the Protocol article's dystopian-sounding title (“Companies are using AI to monitor your mood during sales calls”), my first reaction was: “great, now we will also have AI telling us to smile more”.
Sentiment analysis is not a new concept, there is even an ABAP SDK by IBM that can be used to get a sentiment score of any text. (It’s free, does some other cool stuff, and I totally recommend checking it out.)
Modern solutions mentioned in the article go beyond the text and add speech recognition and computer vision to detect emotional attitude. Considering that Alexa still has mixed success recognizing my verbal commands, I wonder if AI would also stumble on my natural facial expression that prompts people to ask what I’m mad about? And will we see a service industry of “poker-face buyers” to counteract any seller AI efforts?
These AI tools can be invaluable in many cases though. For example, one of the challenges of virtual events is that the speakers can’t quickly glance over the room to gauge the participants’ reaction. Same goes for online training: analysis of the learner reaction could be very helpful to the trainer. I see great value in these solutions beyond the eye-rolling “sales call” use case. JP
Worth A Thousand Words
I wake up every day excited to read news of what's next in the world of AI. The day I saw DALL-E 2, like some of my previous awe at GPT-3, stands out as a watershed moment.
DALL-E 2 is deceptively simple. You just feed it a text description of an image you want - and then it makes that image. And not some ugly weird pile-of-melting-eyeballs image like DeepDream - it's photorealistic if you want. It's style transfer if you want - like the above "a painting of a fox sitting in a field at sunrise in the style of Claude Monet". That text is all that DALL-E 2 needed to make the image above.
The features are so amazing. It can do "inpainting" - take an existing image and update it. Give the kitten a police hat! Turn a picture of your cousin into a Warhol!
Why am I geeking out about this in a business-tech newsletter? Think about how much time and effort businesses spend on making and editing images. Think about a Visio diagram appearing just by speaking into your computer microphone. If a picture is worth a thousand words, and communication is the core of business, then surely a tool like this (or its next generation) will have the capability to enhance business. PM