826 views1 Upvote6 Comments

Senior Director CIO Office in Software, 1,001 - 5,000 employees
We’ve all heard the William Gibson quote, "The future is already here, it's just unevenly distributed." One of the areas that fascinates me is machine learning (ML), including the modeling activity, the support in production activity and the data conditioning activity that occurs upstream of the modeling. What's a reference tech stack for somebody that's trying to build a sophisticated ML operation, and how do you get there incrementally?

I've been involved in a lot of companies that have very sophisticated data warehouses and operational reporting. They’ve even had business intelligence capabilities to look at trends and do forecasting at a certain level of detail, but the modern world of data analytics is a completely different universe compared to the normal technology pieces that you would put together. It spans everything from reverse ETL procedures for sourcing data out of a warehouse, to real-time streaming data out of IoT devices, then conditioning that data and feeding it into some modeling routine, building the models, testing the models, etc., and then supporting them in production.
Vice President for Information Technology in Education, 1,001 - 5,000 employees
What I'm most excited about is not predictive analytics, which tell us where things are going, but prescriptive analytics that tell us what action to take and automate the action. That is where all of this is going. In higher education, one of the problems at public two-year schools in the US, is that we only have about a 40% to 45% graduation rate. Students don't persist, so it'd be interesting to look at the data, because we know the characteristics of why students don't persist. I think the secret sauce would be to reach out to the students and say, "We realize you've registered for these two courses, but students who've done that before haven't succeeded. Instead, we think you should take these classes and think about this first, because when students have done that, 85% of them have graduated with a grade of A or better."

When I look at the airlines, the automated systems I like are the ones that do a good job at predicting what I'm likely to want to do next, not just telling me what I've done in the past. We ascribe a lot of human intelligence to so much of what we do, but it's just a matter of looking at data and choosing which path to take. We can automate more of that and get smarter at automating those processes, because we like to think that a lot of what we do is complicated but it's not. It's just decision trees.
VP, Information Technology in Consumer Goods, 10,001+ employees
The things that we're pushing for are around the greenification of IT. We suck up a lot of power through our data centers and computers, so how can we do that in a smarter way and be part of the sustainability cycle that all companies are being pushed to go through? For example, could you use all the heat in the data center to warm the water before you throw it into the water heater?

We do a lot of computational stuff on site as well. We have clusters that run and they suck up a lot of energy, so why not use that for something better? Or if you have roof space, stick some solar panels up there. Then there's the whole discussion around blockchain, not for cryptocurrency, but for smart contracts and security compliance. How do you use it to ensure that stuff that you put through the supply chain hasn't changed?
CISO in Consumer Goods, 201 - 500 employees
Just waiting for the day when innovation actually takes shape in the security MDR space, instead of everyone touting the best AI/ML solution that is just rebranding of their current technology. 
Director of Engineering in Finance (non-banking), 1,001 - 5,000 employees
There is still huge percent of world's population that don't have access to a bank (a.k.a unbanked) and still there is another huge percent that pays very high fees to exchange money with their friends or businesses and it takes much longer, sometimes even weeks to receive money. This is due to highly regulated banking industry which, for right reasons, need to operate the way it is and focus on anti money laundring.  Levering technologies like AI/ML, Data Analaytics etc. innovatively to make it easier for financial organisations to stay on top of regularions/compliance and yet enable people to make money more relatable, personal and immediately available, when they need it.
Senior Solutions Architect in Services (non-Government), 501 - 1,000 employees
IaaS at the speed of innovation:

Given the push by the application development community for faster delivery of IaaS at the velocity of innovation, it becomes increasingly apparent that Infrastructure frameworks need to support three vehicles currently for hosting of these new cloud native applications with a fourth on the horizon. The three primary delivery vehicles are currently:

- Virtual Machines
- Baremetal
- Containers

The fourth category of delivery mechanism about to be made available to the mainstream is a new methodology known as:

- Serverless

The term "serverless computing" is truly a misnomer referring to a cloud-computing execution model in which the cloud provider runs the server, and dynamically manages the allocation of machine resources. Cost to the consumer is based on the actual amount of resources consumed by an application, rather than on pre-allocated units of capacity. Mirantis believes that this new allocation methodology will become a prevalent way of utilizing on premises resources in the most efficient and cost effective way.

Content you might like

Director of IT, Self-employed
Enabling GenerativeAI in our detect product.
Read More Comments
1.6k views1 Upvote2 Comments

Director of IT in Education, 10,001+ employees
Learning, Pseudocode, Code completion, quick answers

1.3k views1 Upvote1 Comment