When it comes to Data and AI, migration to cloud is good, but thinking out of box is better!
In past 10 years, more and more companies taking their workloads into cloud, in other words they are “migrating”, which is great. Tones of flexibilities, scaling, cost reduction, innovations and much more benefits you can name it from these migrations. And even sometimes they migrate from on-prem to cloud just to modernize their technological environment, which is also great.
Although there are plenty of good reasons to move to cloud, when it comes to Data & AI, before you decide to close your eyes and say, “lets migrate”, think out of box. What do I mean? You probably have seen lots of companies who are surrounded by lots of traditional legacies systems, outdated systems, huge warehouses, old big traditional databases etc., right? Correct, migrating all these legacies can take years and tones of time and effort. Is it really the only way we can modernize our systems and make use of cloud technological environment?
Please do not get me wrong, I am not against migrating on-prem to Cloud, quite in reverse. New technologies have mapped lots of practical solutions to our daily requirements which otherwise in on-prem we would spend years to get to those flexibilities. ?I believe though, if the legacies are stopping you to innovate, to accelerate technological modernization in order to have less cost and more revenue, then THINK OUT OF BOX.
What do you need for your line of business to be more productive? How can you enable your business area to spend less time in old tasks and think more innovative? Is there any way to modernize the environment and build insightful information for them without “migrating” all the legacies?
There are plenty of solutions out there to help you building a complete analytical portfolio for Data Lake / warehouse migrations: from Data Factory, dataflow to Databricks and Synapse analytics, from traditional data warehouses to modern ones, from visualization services to ML & AI services. And you would be able to take all your legacies and start building your insights over those data points. however, you might think of building new brand solutions. For example, imagine your website and all your mobile apps are sending data to your legacy databases for decades and for a new business requirement, you need to build real-time decision-making system, or even you need to build an interactive near real-time analytics platform, where your team can make faster decisions.
In these scenarios, to address those requirements, in many cases, companies will have difficulty to build modern technologies over legacies and consequently they decide to “Migrate” to cloud. Then TIME and COST would be your both big blockers to easily and rapidly get adapted with new technologies that in recent years were built to come helping you. You might miss it. Legacy could always push us to stay behind. Do not let this happen.
?
Instead, you need to think of a way to be less dependent on your legacies over time. As the result, with less and less dependencies you need to migrate just that portion of your legacies as you really need.
Send your data into new environments also / Build cutting edge solution from starch / use prepared AI services and reuse them in your benefits...
领英推荐
3. There are plenty of prepared AI services, already ready for you: speech recognition solution, image detection APIs, sentiment analysis AI services, recommendation services power by AI, … USE THEM.
What is the Azure time series Insight?
Azure time series insight is an IOT type of solution. The IoT landscape is diverse with customers spanning a variety of industry segments including manufacturing, automotive, energy, utilities, smart buildings, and consulting. Across this broad range of industrial IoT market, cloud-native solutions that provide comprehensive analytics targeted at large-scale IoT data are still evolving.
Azure Time Series Insights Gen2 addresses this market need by providing a turnkey, end-to-end IoT analytics solution with rich semantic modeling for contextualization of time series data, asset-based insights, and best-in-class user experience for discovery, trending, anomaly detection and operational intelligence.
A rich operational analytics platform combined with our interactive data exploration capabilities; you can use Azure Time Series Insights Gen2 to derive more value out of data collected from IoT assets.
?
Data Duplication Cost effectiveness versus full migration
In some cases, duplicating some portion of data source is much more cost effective than fully data source migrations. Especially if your solution for LOB requirements generates clear profitability. ?We should not probably leave the profit because of those legacies, correct?
I fully understand that in some cases, full migration is the best and the only option on the table. And thinking out of box might seem na?ve, but if there is anyway, think out of box could help you prioritize your migrations based on two important variables of cost and time, then you might define a better migration strategy and not letting the nightmares of complexity of those legacies take your night sleep away.
Thank you!