Impact of AI on a digital company’s carbon footprint — Richard Whitehead // Moogsoft Inc.
Richard Whitehead
Moogsoft Inc.
- Part 1Application of AI techniques to IT Ops — Richard Whitehead // Moogsoft Inc.
- Part 2 Impact of AI on a digital company’s carbon footprint — Richard Whitehead // Moogsoft Inc.
Show Notes
-
02:22The CTOs focus on observability and why it mattersObservability is essentially an extension of traditional monitoring that gives you a holistic view of whatever application or digital experience you're trying to deliver
-
03:33In this world of clustering and using NLP to improve models instead of adjusting code, how can CTOs use AI to spot trends that are revealed in the matrix of the log files but are not ostensibly visible when they're doing filtering?One of the techniques we're seeing in the industry is the analysis of logs to look for particular messages. Trying to pull more information and extract more value from a large log file is something we're applying machine learning to solve as an industry.
-
06:31The challenge of training modelsMoogsoft does not have pre-trained models; instead, they determine the appropriate algorithm to apply to the customer's data.
-
11:15The environmental cost of storing large dataCompanies are concerned not just about the cost and the financial implications of storing data but the environmental implications of storing large amounts of data. It changes the conversation of what is relevant to store.
-
15:08The log files, what do you do with the data? Should you store data just in case considering the cost of storing data?A paradigm shift is to say no. That data is learning data and creating models, validating the models, supervising those models, and then you store kilobytes instead of good gigabytes.
Quotes
-
"Observability is essentially an extension of traditional monitoring that gives you a holistic view of whatever application or digital experience you're trying to deliver. Very naively, what observability is, it's the addition of tracing to traditional monitoring and log analysis. - Richard Whitehead
-
"One of the techniques that we're seeing in the industry is the analysis of logs to look for particular messages. And the individual message in a log file may not be useful. Generally speaking, log messages are very poor-quality sources, and people tend to only look at them when trying to diagnose a problem." - Richard Whitehead
-
"Trying to pull more information and extract more value out of a large log file is something that we're applying machine learning to solve as an industry. The same applies to anything condensed to a rate." - Richard Whitehead
-
"The problem we encounter at the level of an IT operations standpoint is that no two infrastructures are ever the same. Now it's possible; you can see repeating patterns and infrastructure, and we will encounter Kubernetes running on Amazon es. That's a reasonably common pattern; many people deploy their code using that pattern. Once that code has been deployed, the results are very different." - Richard Whitehead
-
"So most of our technology, and yes, we have a data science team, what they're doing is determining the appropriate algorithm to apply to the customer's data." - Richard Whitehead
-
"It's an unsupervised technology, so you just start running, the customer starts sending the data, and the algorithms are going to operate on the data it sees." - Richard Whitehead
-
"A common question is, can Moogsoft take the model information from an existing customer? And apply that to the product so that other product users can leverage that information. And the answer is no, for my reasons, which are your environment may look the same, and you're an airline in Europe with the same technology. But no, your data is completely different because you use a different application." - Richard Whitehead
-
"In more recent times, that conversation has turned not just to the cost and the financial implications of storing data but the environmental implications of storing large amounts of data. And, as is common with so many things, this is becoming front of mind for large organizations due to legislation." - Richard Whitehead
-
"When I look at some of my customers' environments and the technologies they use in the machine-to-machine interface, like maybe using JSON payloads or even hopefully not XML because that's horrible, this is my opinion. But shockingly, I'm amazed by how many systems use email as a machine-to-machine communication mechanism." - Richard Whitehead
- Part 1Application of AI techniques to IT Ops — Richard Whitehead // Moogsoft Inc.
- Part 2 Impact of AI on a digital company’s carbon footprint — Richard Whitehead // Moogsoft Inc.
Up Next:
-
Part 1Application of AI techniques to IT Ops — Richard Whitehead // Moogsoft Inc.
What is the world of artificial intelligence inside IT operations? There is the creation of models based on event data, log data, and whatever interests machines and DevOps. And then there is the ability to simplify or automate some, whether it's an alert or seeing something in the data that the operator might not see. What more? Listen to Richard Whitehead, CTO at Moogsoft Inc., talk about how artificial intelligence will transform the IT industry by applying AI techniques to IT ops.
Play Podcast -
Part 2Impact of AI on a digital company’s carbon footprint — Richard Whitehead // Moogsoft Inc.
Digital companies generate tons of data. The problem is no longer the cost of storing extensive data but the carbon footprint it generates. With new regulations coming for companies to report their carbon footprint and financial reports, what's the smart way to save on data storage? How can AI and machine learning help? Listen to Richard Whitehead, CTO at Moogsoft Inc., talk about the impact of AI on digital companies' carbon footprints.