Among all the novel new applications of analytics that are making rounds in the enterprise these days, few are generating more buzz than chatbots. Last week saw Facebook Inc. add fuel to the fire by introducing a new machine learning tool aimed at letting developers incorporate the technology into their projects more easily.

Bot Engine, as the framework is appropriately called, provides the ability to create automated question answering functions that can adapt themselves to user preferences. The starting point is a set of interaction rules defined by the developer that provide a basic guideline for how to handle inquiries. As an application starts amassing chat logs, it’s able to take advantage of the tool’s machine learning component to scan the information for recurring communications patterns and factor them into its programming. According to Facebook, the approach helps improve both the accuracy and variety of responses over time.

A similar value proposition is being put forth by DigitalGenius Inc., a chatbot startup that raised $4.1 million in funding last week to harness the technology for automating customer service operations. Its namesake platform works similarly to Facebook’s new tool and uses an organization’s contact center logs to customize a bot’s behavior according to  user behavior patterns. The outfit says that the software can thereby kill two birds with one stone and reduce help desk staffing requirements while enabling consumers to have their questions answered faster.

Yet as useful as they are, chatbots represent only the tip of the machine learning iceberg. To help organizations explore the other uses of the technology, Google Inc. last week released a new iteration of its TensorFlow artificial intelligence tool that can scale across upwards of thousands of machines. The update enables the system to support much larger and more sophisticated projects than it could before. 

Image via blickpixel