ML jobs can be divided in two sects :
The experimentation phase involves the steps where you are trying to build the ML model for the first time for a particular use case. Now this takes time to understand the presented data, have to and from discussions with stakeholders, understanding their requirements and then putting the data through pre- processing to feed into the ML models.
You have to keep your observations noted of what can be done with the existing data and what cannot be done, which can help in achieve the objectives or you need to take a detour in between.
The MLOps phase comes in when the experimentation phase is complete and you are sure of your data, features and models that they are working well and now it needs to be put into production to automate the whole ML process.
Either you can get both the scenarios above or you can get projects where you are in a support role for an automated ML pipeline to monitor and present the findings after every run, usually after every week.
But one thing is for sure, it's not just training in the companies that is being done. It's a lot more than that and you either need to adopt MLOps practice or get into data engineering as per today's market trends since just keeping models into the notebook is of no use.
Any ML/AI engineer who can share, what does a day to day work look like for you ? Is it mostly a research based role ? Do you code regularly ?
Hi, I am a software engineer with 12 years of experience. I need some suggestions. 1: Is it too late now to learn Data Science and pursue a career in it? 2: If I want to learn the necessary concepts of ML what path I should follow? 3: I...
Hi everyone,
I'm currently working on an SAP MM-based project. Despite talking with everyone who could assist me in securing a development role, I've been told that, as an ASE, I can't switch to development. I'm worried about my career ...
