Why Wait? Use AI to Automate Basic Business Tasks

Print Friendly, PDF & Email

In this special guest feature, Brian Schwarz, VP of Product Management at Pure Storage, discusses how the primary value of AI is around automation. That means just about anywhere we currently use humans to complete repetitive tasks, AI can help. AI and machine learning teach computers to recognize patterns in unstructured data and turn it into structured data in a manner that allows for automatic response to be applied. Industry analysts estimate AI will be built into nearly every new software update to hit the market by 2020. Organizations that make the shift earlier will have a major competitive advantage. This article will explore how AI can boost your business, implementation measures and how data can guide this transition. As VP of Product Management for Pure Storage, Brian is an integral piece and early member of the team that created FlashBlade. He brings a wealth of experience in data center infrastructure and an expert on the intersection of AI and storage, with an emphasis on data-intensive and emerging workloads. Prior to Pure, Brian spent time at Cisco as well as Symantec, which he joined via the company’s merger with Veritas Software.

The primary value of artificial intelligence is around automation. That means just about anywhere we currently use humans to complete repetitive tasks, AI can help.

AI and machine learning teach computers to recognize patterns in unstructured data and turn it into structured data in a manner that allows for automatic response to be applied. There are near countless examples – HD cameras in a factory or warehouse used for image recognition to improve quality control or to perform inventory management, retail vendors using video to recognize and process customer behaviors, even image recognition in the medical field for pre-screening X-Ray and MRI images will be impacted by AI.

The insurance industry, as another example, uses AI to help with estimate analysis on claims: A picture of the a crashed car or other property damage can be run in near real-time against the company’s database of damage photos to give an instant estimate for a claim payout, all without having to send a person out to assess the damage.

Industry analysts estimate AI will be built into nearly every new software update to hit the market by 2020. Organizations that make the shift earlier will have a major competitive advantage.

Don’t Fear the Robots: How AI Can Boost your Business

Many of the tasks AI replaces previously required people resources that are now available for redeployment to more critical, strategic parts of the business. There are significant cost savings to be had, but the “robots are going to automate us out of jobs” story is over-played.

The real opportunity is to unlock new opportunities and concepts that weren’t economically viable in the past. AI enables new projects and innovations that were previously thought to be too costly or time-consuming. The new project isn’t cost effective to move forward if the cost is 100% manual, but if the cost can be reduced by 50% through the use of the AI the project now becomes possible. By harnessing AI, creative innovators will open a wider range of opportunities. An executive or IT professional deciding when and how to put AI to work should stop to consider which valuable projects they avoid due to expensive, repetitive tasks that drive up cost.

Areas where AI can assist with automation include customer support (sentiment analysis), manufacturing and QA. Job interviewing is another emerging use case – using video processing to make the process of identifying a candidate’s intangibles more scientific. It turns out that certain facial expressions and other human behaviors are repetitively successful in sales situations; the same is likely true for customer services and other interactive aspects as well. AI helps us identify those patterns more readily. The creation, implementation and ongoing use of these technologies takes people. A recent survey by CapGemeni asserted that AI will create more jobs than it will destroy. Much like previous forms of technology have created more value than they eliminated.

Putting AI to use to automate basic business tasks can result in better quality of service, better customer experience, better project implementation and lower cost. As an example, autonomous, 18-wheel cargo trucks will save money on salaried drivers, but there is additional value in a truck that can drive cross country, 24 hours a day, seven days a week without stopping for food or sleep. In shipping, time to delivery is an incredible business value, but it isn’t cost effective to fly everything. Perishable food shipped on an autonomous vehicle might see two-to-three additional days on the store shelves, which reduces waste and improves quality.

Crawl, Walk, Run To Get Started

Gartner says data centers that fail to apply AI and ML will cease to be operationally and economically viable by 2020. This isn’t science-fiction technology to be considered in abstract terms – it’s ready to go now, and it’s critical to the future viability of nearly any organization regardless of industry.

In the near term, IT won’t necessarily be driving the business toward AI, but rather the other way around. Line of business owners will have needs that drive IT to use AI. At many organizations, AI might start as a one-off research project outside of “traditional IT.” As those projects succeed and the cost and simplicity of AI comes down, the projects will move to IT for scale and integration with the rest of the structured data processing that takes place within IT.

That’s because while AI often processes unstructured data – video, images, audio, etc. – it produces structured data that needs to be integrated with other structured data to perform a complete solution. Consider an example of an automated “greeter” at your local hardware store. A customer can walk in, hold up the nut or bolt or whatever they are looking for in front of a camera and it will tell you which aisle the object is and whether it’s in stock. The image recognition needs AI-enabled training, but the system needs to integrate with the product catalog (new item introduction) and the inventory control and POS systems as well.

One way to get started is with a small, executive-sponsored R&D effort outside of normal operating groups. Seed a few ideas but give them wide ability to innovate and prototype a few solutions. Review and see if one can be pushed forward into production. Have the team repeat this process until a few successes exist before trying to mainstream the results.

Let the Data Be Your Guide

First and foremost, organizations need to decide where they’re going to run AI tasks – public cloud or private cloud. Let the data be your guide. In AI, large data sets win when it comes to training AI neural networks, so organizations have to do the AI work where their data exists. Data has gravity – it is too expensive, time consuming, and complex to move large data sets, so the project has to move to the data, not the other way around. If the data is on-prem, and will be integrated into other on-prem structured datasets then it makes sense to run your AI projects there as well If the datasets are generated in the public cloud as part of a web services, or as the natural landing place for you IoT data it makes sense to run your AI data there.

All industries – without exception – are going to be impacted by AI, but why and how isn’t entirely predictable.

The winners won’t necessarily be the biggest companies or the ones with the best marketing, but rather those that are most adaptable and innovative. The leaders in AI will re-shape the way business is conducted – in some senses it has leveled the playing field and give companies of all shapes and sizes new opportunities to innovate. Carpe Diem!

 

Sign up for the free insideBIGDATA newsletter.

Speak Your Mind

*