We got burnt twice on Purview implementations (first time was when it first came out, second was when "we fixed the problems and revamped") awhile ago. I highly doubt my company would even consider it no matter how much I asked. The column level lineage was one of the most important features to me when they rolled it out initially and they don't even have that now?

@Udayan Raju Yes, and because you have to enable all tables, you can easily go over your dataverse storage limit, resulting in significant costs.

Lots of customers in the preview programs are reporting that for D365 ERP, fabric link is not fully baked, does not update as fast as the Synapse Link with Fabric option, and the additional charges for having a full replica of your data in the dataverse (which is required for Fabric link) is quite high. I've frequently seen others report having to redo their entire setup when something goes wrong with a data flow issue. MS seems to be coming out with an "archival" option to potentially address this [my take on it after meeting with MS], but there are a lot of delete issues downstream that don't seem like they've been fully thought out yet. Conversely, setting up synapse link with any network option that isn't public is a MAJOR pain that even Microsoft has trouble supporting. And the archival feature (as it was presented to us) will definitely cause issues here.

I'm a Data Architect (and Data Engineer) for our Data Analytics team and I would suggest you take it. D365 can be quite complicated under the covers and many companies want to use the data down stream for a data warehouse and analytics. But there are very few people (even at Microsoft) that have experience of getting things from D365 (in the data verse) in to Synapse or Fabric. I work with D365 F&O, which is similar to Business Central (BC is a bit lighter from what I understand). It's not a typical normalized database. Here are my answers to your questions. Questions for the Community 1. Is Dynamics 365/Power Platform experience valuable for a future in data analytics? Could skills in Power BI and Power Automate transfer well if I move into a Fabric-focused role? --100% yes. PBI is a part of Fabric and was pre-Fabric until they roled Synapse into it. Power Automate would also give you familiarity with data flows and pipelines that you would use in Fabric and/or as a data engineer. 2. What specific skills or tools will I likely gain from this internship? How relevant are these skills for a future data analytics position? --Building models in PBI and understanding how to design semantic models well is arguably a core skill needed by many data professionals. D365 resides in the dataverse and take a special skill set to connect to Fabric/PBI. You've got some really great data engineering skills covered there. Even if you don't build out the solution, learning all the components will give you a step up. 3. Are there opportunities to develop data analytics skills within this role? For instance, will I have the chance to work on real data analysis projects, or will it be more focused on implementation and automation? ---You have to walk before you run. Sounds like you will get a wide range of skills at the intern level, including some analytics with PBI. Once you develop that skill set, you would have a great foundation for producing something on your own of value to the company that would make you stand out. 4. Transitioning from ERP-focused roles to analytics: Has anyone here moved from ERP/Power Platform work to data analytics, and if so, what was the learning curve like? 5. Should I wait for a more analytics-centered role? Given my long-term goals, would it make sense to keep searching, or could this internship still be a strong stepping stone? --I'm having to learn the ERP side as someone coming from a long time data analytics role. I've lead data analytics teams and to me the ERP would def be a huge bonus if I was looking to hire a junior person on my analytics team. Just be aware that both the ERP and the Fabric/Synapse side of the fence requires a lot of learning that is always ongoing. If that isn't your jam, than the data engineering/analytics role probably wouldn't be a good fit. For an intern role, this would be a very strong stepping stone in my opinion (remember intern, junior, and often even senior roles are mostly not all fun stuff. A really high percentage is just getting data into good shape or finding out how architecture components don't play nice with each other. )

@Anko Top That's a really good point about moving workspaces if you find your original plan doesn't work. My head keeps getting stuck on the what-if use cases and that tidbit makes it easier for me to decide how to set it up. thnx

The short answer is Synapse Link is more complicated to implement with more limitations from a security standpoint while the Fabric link can be more expensive (because of charges on the Dataverse side once you go over a limit). I've heard that they are looking to address this in the future.