[throwawayge]

As a GE SW developer, I could chime in.

Regarding digital transformation, this should not come as surprise but there is an incredible tension and misalignment between a GE core pushing for a platform and different businesses.

The issue is that businesses have roadmaps where they need to release new products each year, and it does not take into account having to rewrite everything or huge part from scratch to get some longer term platform benefits. So in the end the adoption is an issue.

I joined GE just before Predix introduction, and no one in my business rushed to adopt it. As a matter of fact, now that GE digital is going to be divested, that was likely the smart move.

Moreover, the product I am working on, is based on the "GE platform" effort of a few years prior. Platform that has no longer any official support (the maintenance team is one guy somewhere in the world) and that has re fragmented in every different product it was used in. The lack of support is not really an issue because the framework has been battle tested but on the other hand, there is no obvious benefits coming from having an "almost-common" core between applications or devices. So as business it makes you wary to adopt the last idea coming from GE digital or GE central, event it makes absolute sense on paper, because you have no idea how long they will commit to that idea. This is really a chicken-egg problem, because no one wants to use a zombie platform, and no one wants to spend support on a platform no one uses.

Maybe on a more positive note, even if I have not seen much direct impact on GE digital products, GE SW culture has changed. When I joined, SW was kinda neglected, more seen as a liability than anything, and now at last we have moved to more modern tools with CI pipelines, focus on devops and so on. But this is more like a grassroots effort. But I don't feel anymore that we are using a sw stack and tools totally outdated.

[gedigi_YB7ZCE5]

I'm a fellow GE developer working under one of the businesses, and I agree with what you've said here.

The team I work on was never sold on the benefits of the platform, but we were told to use it. Arbitrary targets like "get 150 apps to run on Predix" were set, but because there were no clear advantages to doing so, these targets were internally de-prioritized compared to working on features. They spun the numbers to give the appearance of momentum, but over time people talked about the platform less and less, and it became obvious that the emperor had no clothes.

GE has pockets of good developers; the grassroots culture that you mention is evidence of that. I'm hopeful that Flannery's renewed interest in the business verticals will enable those developers to focus their effort in places that actually deliver value, instead of wasting it on grand top-down initiatives to become some kind of industrial AWS. The sooner that happens, the sooner the company will get out of the miserable slump that it's been in.

[le-mark]

How's the culture? Are devs actually busy and productive or is the the typical never ending wait for requirements to materialize that's so common at moribund large enterprises?

[throwawayge]

I certainly cannot speak for all teams, but most teams are fairly autonomous as long as they deliver. Like any big company, there are lots of stakeholders so that means that final decisions, for example a feature being in or out of scope, can take an unnecessary long time, but it does not mean you are stuck waiting. There are always lots of things to implement, refactor or improve that do not require a seal of approval.