Most people feel a bit astounded when they type a query into Google and get relevant results in milliseconds. They’re probably not as impressed when using an enterprise search feature at their workplaces to find out about the help desk hours or which paperwork to file for maternity leave.
It would seem that working with a relatively limited amount of information would increase the likelihood of getting helpful results. The reality is, though, that the probability of a person getting the answers they want or need goes up with the amount of available data. Enterprise search is not a new option, but companies have struggled for years to improve it.
In addition to the issues associated with a limited pool of data, organizations often deal with information silos. Creating a search interface for employees, then, usually means determining the origin of information, which may mean hunting for it by perusing dozens of tools and interfaces. Do solutions to these problems exist? Amazon is confident that machine learning and natural language processing (NLP) can help.
Amazon hopes to tackle enterprise search with Kendra. It’s a tool that offers suggested answers above the main group of results, much like Google does with its Featured Snippets. Since the product uses natural language processing, people can ask what they need to know by inputting whole phrases rather than a few keywords.
Amazon Kendra also has machine learning algorithms that get smarter over time. They learn which answers users find most valuable, thanks to a quick click that lets a person confirm whether the information provided was useful to them or not. Processes are as easy from the perspective of those tasked with uploading a company’s data to Kendra, too, thanks to a simplified approach that enables speedily pulling in information from multiple sources.
Knowing the basics of how Amazon Kendra functions should make it easy for people to understand why the tool could be a game-changer for helping employees get prompt, accurate information. Similarly, organizational leaders will see that artificial intelligence (AI) and machine learning algorithms are not out-of-reach technologies. A broader effect of Amazon Kendra, then, could be that it makes companies more curious and eager to learn about AI’s capabilities.
Moreover, opportunities exist for companies to use Kendra to assist their customers. When Amazon published a recent press release about the general availability of the tool, they mentioned several case studies of existing customers. One of them is PwC, which has 276,000 employees in more than 150 countries. It put Amazon Kendra’s machine learning power to use by creating RegRanger, which gives clients in regulated industries reliable information.
Kendra is part of an overall trend whereby companies have more opportunities to implement AI in practical ways to meet goals. Chatbots are prime examples, especially since they integrate seamlessly into a current workflow and complement human efforts. Chatbots engage with customers over multiple channels, too.
Once tools like Amazon Kendra show enterprises what’s possible with help from machine learning and NLP, more companies may feel encouraged to explore using them. If that happens, the businesses will see that AI is accessible and able to assist with particular challenges.
Another reason why Amazon Kendra could unlock the potential for businesses to use is that it can handle queries containing highly specific terms, from domains such as financial services, pharmaceutical or energy. Thus, if a business leader feels interested in using AI but believes the language of the industry is too advanced for a machine to understand, that may not be true.
Amazon Kendra could be the tech tool that proves how innovations are progressing so much that most industries will find they’re worthwhile. A company merely must connect its data repositories to the search engine to make the tool get to work in finding the answers people need to know.
Amazon Kendra aims to demonstrate that getting useful information in seconds while at work is not an impossible goal. Since the tool understands industry-specific terminology, business representatives can look forward to it meeting expectations from the start, rather than engaging in lengthy training sessions.
Plus, the tool could drastically reduce the time taken by human resources executives or department team leaders as they attempt to steer people in the right direction when they seek information. The company as a whole then benefits from the increased productivity from the time saved.
It’ll be interesting to see how many companies end up using Amazon Kendra, and the respective sectors in which they operate. For now, it seeks to solve enterprise search issues and seems likely to succeed.
By Kayla Matthews