by Ralph Tkatchuk

5 ways companies should use AI

Opinion
May 19, 2017
AnalyticsBig DataE-commerce Software

AI is no longer a thing of dystopian sci-fi novels. Today, it has real potential to assist companies and individuals for fast, better and smarter operations

7 artificial intelligence
Credit: Thinkstock

Artificial intelligence (AI) was once a topic reserved for high-level computer scientists and futurists. Today, it doesn’t come with such daunting baggage.

Developments in the field have made AI accessible to just about everyone. AI subfields such as machine learning and natural language processing have even become buzzwords that we now constantly hear and read about in the news. And according to estimates, by 2020, the AI market will approach $50 billion.

+ Also on Network World: Why AI will both increase efficiency and create jobs +

Writing about the expanding AI market, California-based entrepreneur Gurbaksh Chahal says AI will eventually reach every industry, including real-time bidding, biometrics, marketing and speech recognition.

Infrastructure and platform providers such as Google and IBM also now offer access to their AI APIs. And cloud services, third-party developers can now use these APIs to integrate AI into their applications. Enterprises are now aggressively looking into implementing AI projects to improve their business functions.

5 ways companies may consider applying AI

1. Business intelligence

Due to the intense competition in today’s business environments, enterprises should always be on top of issues before these become major problems. Thanks to the sophistication of analytics tools available today, decisions are now driven by data. This way, decision makers can avoid the pitfalls of biases and relying solely on intuition.

As such, there is a constant need to monitor and collect business information from which insights are to be derived. Managing the data becomes a major effort due to the sheer amount of it coming from a multitude of sources. Traditional business intelligence (BI) methods and tools struggle to cope with such volume and variety.

This is where AI comes in. Advanced algorithms are now used to crunch massive amounts of data and generate reports.

AI can even create visualizations and dashboards in real time. AI also can be trained to detect outliers and even monitor for thresholds based on key performance indicators so that the system can send out alerts and appropriate action can be done in a timely manner.

2. Chatbots

Another interesting development in AI is chatbots. Developments in natural language processing has enabled systems to be able to process many conversational prompts. While experts would argue that the Turing Test is far from being beaten, the widespread use of virtual assistants such as Apple’s Siri or Microsoft’s Cortana for routine tasks is just a glimpse of what chatbots can do.

Popular messaging applications now opened up their development platforms to allow others to integrate their bots into these chat apps. For example, users can now order from 1-800-Flowers.com through Facebook Messenger. CNN’s Messenger bot can be asked about the latest events happening around the world. Even Uber allows you to book rides within the chat window.

For businesses, chatbots can provide a level of automation and free up resources used to manage functions account management and routine support. Other applications of chatbots even include ecommerce and online shopping. Users can simply ask the bot about certain products or provide descriptions of what they want, such as “Show me all your red shirts size small,” and the bot will respond with recommendations all within the chat window.

3. Localization

Localization or providing a user experience unique to a particular market has become a major consideration for today’s ecommerce.

Previously, only those engaged in digital goods enjoyed an easier time participating in cross-border ecommerce. However, logistics and payment services have become more reliable that even businesses dealing with physical goods can now easily consider going cross-border.

Merchant services have made pricing, taxation and shipping easier, but language continues to be a barrier to cross-border ecommerce. Imagine having to translate all your product descriptions into at least 10 languages. To address this, translation APIs can be used to handle the translation.

While there’s still much premium for humans to do the translation for content that’s context-appropriate and style-consistent, attempting to do this on thousands of items may not be cost effective or time efficient. Natural language processing AI can now produce straightforward translations with a fair degree of accuracy.

4. Personalization

Aside from localization, another way for businesses to provide a more engaging customer experience is through personalization. Using AI, personalization can now go beyond keeping one’s browsing settings and preferences.

Ecommerce can now provide recommendations based on a user’s browsing and buying history. We’re already seeing this in Amazon’s recommendation. Recommendations are reported to account for 35 percentof Amazon’s sales and is considered one of the company’s key technologies.

Today, businesses can also implement their own recommendation algorithms using machine learning APIs. Businesses can train their algorithms using customer tracking data in order to provide recommendations to other items in their inventory, which could lead to more sales.

5. Automation

Businesses that rely on manufacturing and logistics can further improve their supply-chain management through AI by effectively managing inventory and streamlining operations.

AI can also track supply and demand in the regions a business serves so that it can automatically adjust inventories, moving stocks from low demand markets to high demand markets. AI can also adjust pricing given data.

Even the actual transportation of goods can be improved by AI. AI is used by logistics companies to plot the most efficient routes to physically move inventory. UPS has been using their ORION software for years to optimize the stops their drivers make every single day.

Getting Into AI

Knowing these applications, businesses can now examine for which business functions they would want to use AI.

There are lower barriers to implementing AI projects today. For example, Google Cloud Platform offers translation, natural language processing, prediction, machine learning, and even video intelligence APIs at pay-as-you-go pricing. Extensive documentation also allows most developers to be able to tap into these APIs without having to concern themselves with the nuts and bolts of creating their own AI. These allow projects to get implemented quickly.

What’s exciting is that AI is still a developing field. As it matures, we can only expect more complex capabilities that could further empower businesses and improve the lives of consumers.