Chatbots: a new, customer-oriented communication channel

It is not just about technology: a chatbot is a new, customer-oriented communication channel that uses machine learning tools.

A chatbot is a software capable of interfacing, on the one hand, with the end-user, relying on natural language, and on the other hand, with information systems, with the aim of supporting the company in its various functions using an efficient and innovative approach. But it is not just about technology. A chatbot is a new, customer-oriented communication channel that uses machine learning tools to connect the company with its stakeholders.

A natural language is a language similar to what is commonly used by people. The chatbot can be configured to use either colloquial speech, or a more formal variant. By default, chatbots utilise a textual language. However, some applications support the integration of voice recognition mechanisms, allowing the use of a natural vocal language, both on the input and the output sides.

Within a business, a chatbot can be used to optimise communications and provide a range of services, to manage an information or sales channel, or to configure products. Chatbots can be adopted to various fields of application, the key condition being that an already structured flow is in place, within which the chatbot can function.

Chatbot technology can also provide in-house benefits, allowing employees who need specific information to interface with a single system, available 24/7, which can concurrently support multiple users.


Technology for the user

Technology at the service of the user

From a technological point of view, chatbots contain a fundamental element referred to as natural language understanding or natural language processing. This makes it possible to take a sentence transmitted in the form of natural language and translate it into more structured information which can be used at a computing level. The chatbot therefore functions like an interpreter, translating natural language into computer language and vice versa. It can understand what type of response it must give the user and provides feedback that the latter is able to understand. In order to interpret the message transmitted and to translate it into aggregate information appropriate for use at a computing level, the chatbot uses advanced machine learning techniques.

Chatbots can use three main communication channels: instant messaging platforms, IoT interface devices and personal robots, which are in turn associated with three different levels of interaction complexity. Messaging applications such as Telegram, Messenger and Skype are the most widely used and correspond to the simplest level of complexity as they only rely on the use of a textual language. In devices with IoT interfaces, such as the tools offered by Amazon (Amazon Echo) and Google (Google Home), a higher level of complexity is reached as the textual message becomes vocal and the chatbot is therefore required to convert the information from voice to text and vice versa. At the highest level of complexity, the user receives feedback from the physical robot, which also facilitates an empathic interaction. The robot can carry out the commands it receives, thus enabling interaction with the real environment. The increased ease of use on the user side goes hand in hand with the increased technological complexity of the chatbot software.

2016 was the year of conversational technologies, an evolution that began with Telegram, the first platform to offer a chatbot-based service. Facebook then made further developments to these technologies, implementing them and broadening their use. They were subsequently adopted by an increasing number of platforms. The enormous development of messaging applications has led to the widespread use of these automated tools, making way for the evolution of bots at consumer level.

Why use a chatbot?

Why use a chatbot?

The strength of chatbots is rooted in their simplicity, as they are able to provide an interactive service based on natural language.

Among customers, the main obstacle to the adoption of a chatbot tool is the inherent doubt concerning the actual feasibility of implementing such an advanced system. For this reason, Machine Learning Reply partners with clients to carry out an initial analysis phase, fundamental to understanding the extent to which the chatbot can be integrated into the service offered, all the way to the development and implementation of the overall project. From here, the system can be extended and if it proves to be successful within the company, it can subsequently be integrated with all the corporate systems.

The chatbot technology applies to any market sector, since all companies have a customer service function, a sales offer and employees to manage. For this reason, a chatbot can be thought of as a new communication channel, with the only customer requirement being an innovative thinking approach, open to the possibility of a digital ecosystem to support the growth of the enterprise.

The application of chatbots in business

Currently, chatbots present four main fields of applications...

Product configurator


The chatbot guides the user in configuring a product. This application, for example, has recently been implemented to create a car configurator for a leading car manufacturer, to enable buyers to choose the characteristics of the car in terms of features, engine and customisation, with the aim of offering buyers a car configured according to their specific preferences and needs.

Smart customer support


The chatbot makes it possible to filter users’ initial requests and resolve them quickly and easily, with solutions that are typically provided by identifying the scope of the request and accessing a database with corresponding problem resolutions. The call is forwarded to an operator only in the event that the chatbot is not able to resolve the problem.

Smart sales


The chatbot facilitates a new and interactive product navigation approach, as it is able to present the product lines within a catalogue offering customised possibilities based on feedback received from the customer. Depending on the customer’s requirements, payments can also be made directly though the chatbot channel.

Business to employee


The chatbot can be used by a company for internal communication, providing employees with services such as form or scheduling requests. Again within an HR scope, it is also possible to develop chatbots that are able to carry out a first screening of candidates for open positions in the company. The bots can identify answers to standard questions regarding the candidate’s education background and experience, with increased efficiency for the organisation, as well as being interesting from the user's perspective. With a chatbot, every piece of data is recorded automatically, thus also providing a valuable information collection service.

The added value of Machine Learning Reply

Within this context, offers an approach that starts with the analysis of the application itself, in order to understand, together with the customer, the exact purpose for which the chatbot is being developed, which platform is better to use and what type of user can take advantage of the end solution.

The resulting step-by-step process includes dialogue and user management, arriving at a complete integration with the customer's corporate systems, thanks to the add-on characteristics that make it possible to complement the system with enhanced functionality. In fact, Machine Learning Reply offers additional services that can be integrated into a chatbot to increase its efficiency, such as continuous learning, thanks to which the chatbot continues to learn from the user through its capacity for interaction, or smart analytics, which are designed to capture customer data classifying their tastes and preferences. From a marketing perspective, the bi-directional nature of chatbots facilitate the development of new products or services for needs which are currently not being met. Chatbots can also request customer feedback on company product and service offers or on a particular service.

The competitive advantage of Machine Learning Reply consists of the ecosystem-based approach, which goes beyond the purely technological context to offer customers a channel that best represents the specific brands for the consumer. The ability to design the type of interaction and brand representation is a major advantage that other market players cannot offer. By being part of the Reply Group, the leading digital enterprise in Italy, the Machine Learning Reply solution offers an ideal ecosystem for the end-to-end development of the entire process of the creation and implementation of chatbots. The company, in fact, incorporates the specific expertise relating to processes typically analysed by chatbots, both in terms of customer care and in sales management, together with the ability to ensure a seamless integration with the customer's other corporate platforms, managing each and every process in a synergistic and centralised manner.

Another very important factor is the advanced service customisation capabilities offered by Machine Learning Reply, which thanks to a technology independent design is not tied to a specific vertical system, instead identifying the best solution on the basis of specific customer requirements, analysing the possible applications of the chatbot together with the customer.

Within the Machine Learning Reply workflow, the chatbot integrates with other features, such as the recommendation system. Since the chatbot represents an interface to the user, it gives the company the opportunity to get to know and contact users, offering them personalised services driven by machine learning systems. Moreover, the chatbot is based on bi-directional conversations which have the potential to serve as enablers for other activities, such as requests for information or products. The actions themselves are not carried out by the chatbot, but by other, more advanced systems including data robotics, which are implemented together with the chatbot. While these actions normally require human intervention, within a computing ecosystem they are automated, thanks to the learning ability offered by machine learning software. The human operator is therefore engaged only in the event that a bot is given a complex or very specific request, to which the chatbot is unable to find an answer. Based on this approach, the operator's role therefore requires a higher qualification level. Within this context, the use of chatbots facilitates the re-insourcing of low-profile jobs within the company, which are often outsourced due to the high associated costs. The company consequently sees actual benefits in terms of efficiency as well as cost.