I transferred into my current job a few years ago, and this year, I interviewed two transfer candidates, one of them (not you, I'm pretty sure) a QA engineer who wanted to move into an engineering role, so I might have some thoughts on this situation from the other side:
What advantages does the candidate expect from being an internal transfer, and what can we fairly offer them? To me, an internal transfer has a sizable advantage in that they are a known quantity in many aspects: You can ask them in much more detail about previous work at the company, you can talk to their co-workers, you have a clearer sense of the projects you were involved in, and in some cases you may already have personally worked with them.
Where we DON'T cut the candidates any slack is the technical side of the job, and QA is usually different work from engineering. So the candidates still stand or fall based on the technical interview.
The controversial aspect of Machine Learning jobs in particular, and the situation you described, is credentialism. In many areas of software engineering, credentials are indeed overrated, but Machine Learning IS a very technical field, requiring rather advanced math to truly understand it (rather than just blindly plug together pieces from an ML toolkit), and right now the field is undergoing revolutionary changes, so this is one area where a recent graduate degree really DOES convey a concrete advantage.
It should not be an INSURMOUNTABLE advantage, but you'd have to be a very impressive non-academic candidate. What ML papers have you recently read? How would you plan to apply them in the job you're aspiring to? What contributions have you made to the field?
I was trying to transfer as an engineer to another engineering position with the possibility of growth. These were not pure research positions, but positions with a fair amount of scaffolding making internal tools or productionizing a system alongside a team of existing experts, and so I thought my good performance being an engineer was going to let me move to a place where I could do more or less what I had already been doing, while learning and having room for growth. I had no illusions about being competitive for the actual machine learning _research_, having only taken online classes and done a side project. But there is a lot of schlep in managing data and then getting a model into production I knew I could do because I had done very very similar things successfully.
I'm not saying the people they picked over me are worse though, I'm just sad they couldn't have picked us both :D
What advantages does the candidate expect from being an internal transfer, and what can we fairly offer them? To me, an internal transfer has a sizable advantage in that they are a known quantity in many aspects: You can ask them in much more detail about previous work at the company, you can talk to their co-workers, you have a clearer sense of the projects you were involved in, and in some cases you may already have personally worked with them.
Where we DON'T cut the candidates any slack is the technical side of the job, and QA is usually different work from engineering. So the candidates still stand or fall based on the technical interview.
The controversial aspect of Machine Learning jobs in particular, and the situation you described, is credentialism. In many areas of software engineering, credentials are indeed overrated, but Machine Learning IS a very technical field, requiring rather advanced math to truly understand it (rather than just blindly plug together pieces from an ML toolkit), and right now the field is undergoing revolutionary changes, so this is one area where a recent graduate degree really DOES convey a concrete advantage.
It should not be an INSURMOUNTABLE advantage, but you'd have to be a very impressive non-academic candidate. What ML papers have you recently read? How would you plan to apply them in the job you're aspiring to? What contributions have you made to the field?