Glossary AI

Get acquainted with some new terms.

Actions on Google

The Google Home is a smart speaker device, similar to Amazon Echo, which can listen to your voice commands and help you perform tasks (such as a switch on the lights). These tasks are called Actions on Google.

Adaptive algorithm

An adaptive algorithm is an algorithm that changes its behavior at the time it is run, based on information available and a priori defined reward mechanism (or criterion). Such information could be the story of recently received data, information on the available computational resources or other run-time acquired (or a priori known) information related to the environment in which it operates.


An algorithm is a set of specific mathematical or operational steps used to solve a problem or accomplish a task. Algorithms play a central role in AI, transforming or analysing data. That could mean performing regression analysis, classifying customers, or finding relationships between SKUs. Each of these analytical tasks would require a different algorithm.

Amazon Chime is a modern, fully-managed communications service from AWS that makes it easy for you to communicate with people inside and outside your organisation using voice, video, and chat. With Amazon Chime, online meetings are easier and more efficient, crystal clear audio and high definition video keep you focused on the discussion, and the service is designed to work seamlessly across desktops and mobile devices. Because Amazon Chime is a managed service that runs on AWS, it doesn’t require you to deploy or maintain complex infrastructure and software. And you benefit from the security that comes with a data center and network architecture built to meet the requirements of the most security-sensitive organisations.

Artificial intelligence (AI) accelerator

An AI accelerator is (as of 2016) an emerging class of microprocessor (or coprocessor) designed to accelerate artificial neural networks, machine vision and other machine learning algorithms for robotics, internet of things and other data-intensive or sensor-driven tasks. They are frequently many core designs (mirroring the massively-parallel nature of biological neural networks). They are targeted at practical narrow AI applications, rather than artificial general intelligence research. Many vendor specific terms exist for devices in this space. They are distinct from GPUs (which are commonly used for the same role) in that they lack any fixed function units for graphics, and generally focus on low-precision arithmetic.

Artificial narrow intelligence (ANI)

ANI covers all of the public applications of AI currently available. It's called "narrow" because an ANI system's knowledge is specific to a limited set of topics. The AI that navigates a self-driving car can't also hold a conversation with the driver.

Artificial general intelligence (AGI)

Also called superintelligence, AGI means the system has intelligence across a broad range of subjects. True AGI does not currently exist, but it is what many people think of as AI because of pop-culture icons like Skynet, from the Terminator movie series. When AGI does come online, it would likely change our current frameworks.

Artificial intelligence marketing (AIM)

AIM is a loose categorization of products, technologies, and services that use AI to enhance or automate marketing. This term has been around for a while, but it remains to be seen if it will catch on or be replaced.

Artificial-intelligence platforms:

Google TensorFlow (TF)

The Google Brain team made its secret sauce, TensorFlow, freely available to the public in 2015. TensorFlow is a machine-learning library that powers RankBrain, which handles a good portion of Google searches. Because it's open source, developers can use and modify its algorithms to build and train their own neural networks.

Watson may be the most recognisable brand name in commercial AI, having famously won the quiz show Jeopardy! in 2011 against past champions. Under Armour's UA Record app uses Watson to power what it calls a cognitive coaching system, from recommending run paths and race preparation methods to helping shape an athlete's meal plan.

Salesforce introduced Einstein in 2016; it is already built into its CRM product. Einstein, it says, is designed to take over mundane data entry tasks and enable sales representatives to better focus on their customers.

Artificially generated content (destined to inherit the AGC acronym)

Artificially generated content is a burgeoning field that relies on natural language generation (NLG) technology. The Associated Press uses an NLG product called Wordsmith to automatically write simple news stories, and Credit Suisse uses Narrative Science to generate commentary on research reports. Fast-maturing match applications of artificially generated content include the mass customization of communications based on specific details about each customer.

API (Application Program Interface)

A set of functions and procedures that allow the creation of applications which access the features or data of an operating system, application, or other services.


Any software that automates a task. In SnatchBot, often used as a shorthand for a chatbot.

Big Data / Smart Data

As the Internet becomes a large part of our daily lives, vast data sets about our online behaviour are available on computer servers for analysis. This is the essence of Big Data, and it represents a big challenge for companies. Analysing this significant amount of data only to keep the most relevant information for a brand has a name: Smart Data.

Brain Technology

Brain technology, or self-learning know-how systems, define a technology that employs latest findings in neuroscience. The term was first introduced by the Artificial Intelligence Laboratory in Zurich, Switzerland, in the context of the Robot project. Brain Technology can be employed in robots, know-how management systems and any other application with self-learning capabilities. In particular, Brain Technology applications allow the visualisation of the underlying learning architecture often coined as “know-how maps”.

Bot platform

A bot platform, which can be found in a chat app, lets developers create bots that offer services, utilities, games, entertainment, and more.

Chat app

An app, like SnatchApp, that enables conversations and messaging.

Chat QR code

Chat codes let you instantly interact with the world around you. You can scan a code with a chat app’s scanner to connect with a new friend or start a conversation with a bot.


Bots are programs that interact directly with customers via natural language processing (NLP, see below). Many companies are already using chatbots for customer support, but the category is loosely defined, and new applications are cropping up. Marketers predict that they will embrace chatbots as enthusiastically as they did apps circa 2013. However, Microsoft's experimental Twitter-based chatbot, Tay, serves as a cautionary tale. Tay made news in early 2016 when Twitter users quickly trained it to spew hate speech.

Cisco Spark

Cisco Spark is an app-centric, cloud-based service that provides a complete collaboration suite for teams to create, meet, message, call, whiteboard, and share, regardless of whether they're together or apart—in one continuous work stream before, during, and after meetings. It is built to help teams work seamlessly.


The Conversational User Interface is the means for dialogue between a Chatbot and a person. The dialogue maybe through text or voice. An example is within a Messaging Platform the text bubble between people having a conversation is the same CUI used by Chatbots. Another example is Google Home which simply uses voice for the conversation. Chat apps provide conversational interfaces that can also serve as computing interfaces. For instance, instead of tapping or clicking buttons to complete tasks, you chat with a bot, sending it commands about the things you want to do.

Computer Vision

Computer vision is a field of technology that lets computers understand what they are seeing. With computer vision, apps can recognise and modify faces, as seen in Snapchat filters. It helps drones avoid obstacles. It helps self-driving cars navigate traffic. Manufacturers have used computer vision for many years for quality control. Now, as the technology is getting smarter, retailers are using it for merchandising purposes, such as promoting a retail kiosk to generate a coupon for an item a customer is examining on a nearby shelf. Facial recognition also has lots of potential for customer service applications.

Conversational Bot

A conversational bot is a chat bot which can carry out conversations which are both continuous and asynchronous.

Conversational Commerce

Conversational Commerce is an approach that puts conversations between brands and their customers at the heart of the online customer journey. Conversational Commerce includes all conversations related to customer engagement and conversion: a chat on a website or on mobile; an interaction on Twitter, Facebook, or Instagram involving the sharing of a link; an offer or the sending of a buy button; etc.

Conversational Intelligence

Conversational Intelligence is defined by a brand’s interaction with their customers on their preferred channels to address their needs. To gain optimal conversational intelligence, you need to get the most valuable information on your CRM. This way, conversational intelligence enables you to personalise every answer you give to your customers. It makes your brand relevant on all channels (physical and digital) and takes into account social media interactions.

Corpus (plural: corpora)

A corpus is the body of text, images, or sounds used to "train" a neural network (giving it labeled examples from which to learn). From the perspective of the brand marketer, this may be the most important input we contribute to the neural network, as it shapes what the AI platform learns.


Microsoft personal assistant software.


Crowdsourcing enables companies to mobilise a group of people to participate and collaborate on a project. This new form of collaborative work gives access to a superior collective intelligence and cost-effective and innovative ideas.

Customer Feedback

Customer feedback relies on the acquisition of information from customers, about a product, the quality of a brand’s service, etc. The objective is to have a representative set of data about customers and analyse it to optimise your product and/or services.

Customer Support Chatbots

It allows the customer to take the question “as is” and put it into a message, then receive a response instantaneously. A combination of Artificial Intelligence and rules-based Natural Language Understanding techniques can attempt to interpret the question and answer it automatically. If the answer to a question requires the system to ask a few questions first (such as “what is your order number?”), it can engage in a short conversation with the customer before looking up the answer from the enterprise backend system. If the bot cannot successfully discern user intent, or answer the customer’s question, it can “transfer” the dialog to a customer service representative in the Contact Center (traditional call center or Customer Engagement Center), either by continuing the chat in the messaging system of the customer’s choice, or moving it over to the voice channel, e.g. by means of a callback.
Applied to customer service, chatbots are also known as Interactive Text Response, a text-based version of well-known Interactive Voice Response, with all of its value propositions but a better consumer appeal, tapping into the 1+ billion users of services such as Facebook Messenger.

Data efficiency

Data efficiency refers to a set of techniques supporting the storage of huge amounts of data. This is an important concept for marketers because many systems based on machine learning require vast amounts of data. If you cannot supply the necessary volume of data, the conclusions drawn from the data are unlikely to be correct. A common example of this is in health care, where systems trained with machine learning are unlikely to be able to diagnose rare illnesses for which large amounts of data are unavailable.

Deep learning

Deep learning is a type of machine learning. Deep-learning systems use multiple layers of calculation. The first layers look at very simple features (lines in an image, for example) while the later layers abstract more complex features (such as faces). Compared to a classical computer program, this is somewhat more like the way the human brain works, and you will often see deep learning associated with neural networks, which refers to a combination of hardware and software that can perform a brain-style calculation. It’s most logical to use deep learning to very large, complex problems.

Directory (for bots)

Bot directories are simply listings of bots, similar to Yahoo’s directories when they first started. It would be reasonable to expect that over time, people are more likely to use bot search engines and not bot directories, but it is still early days. BotList and BotPages are two of the more popular bot directories today.

Ecosystem (Technology Definition)

Technology ecosystems are product platforms defined by core components made by the platform owner and complemented by applications made by autonomous companies in the periphery. These ecosystems offer solutions comprising a larger system of use than the original platform owner created and solve important technical problems within in the industry. In successful technology ecosystems, it is easy to connect to or build upon the core solution to expand the system of use and allow new and even unanticipated end users. The core firm’s product has important but limited value when used alone but substantially increases in value when used with the complementary applications. Technology ecosystems include well-known smartphone platforms, such as Apple and Android, but are also common in gaming consuls and social media platforms. They exist in industrial sectors, where core products in software, manufacturing or scientific machinery nourish an extended community of service organisations that operate as semi-autonomous value-added resellers. (Source: Financial Times)


Wikipedia’s definition of a feature is good: “an individual measurable property of a phenomenon being observed. Choosing informative, discriminating, and independent features is a crucial step for effective algorithms.” So features are elements or dimensions of your data set. For example, features in a set of customer data might include demographics such as age, location, job status, or title, and behaviours such as previous purchases, email newsletter subscriptions, or various dimensions of website engagement.

Full-stack platform

Often associated with the developer profession, the term “Full-stack” has different meanings according to the companies using it. But the main characteristic of these companies relies on their ability to provide and control the service and experience they are selling. Systematically, the “Full-stack” approach implies extending the fundamental skills of the company beyond the traditional marketing, commercial and technological skills.
Beyond mastering the production chain, the combination of technology with business innovation is characteristic to full-stack. In some cases, the experience implies moving from the digital world to the world of atoms: Tesla manufactures its own cars, Netflix produces television programs, etc. In other cases, there is no need to manipulate tangible assets but you still have to combine perspective technology with functional, service-based and operational expertise.

Graph analysis

In the machine-learning context, a graph can represent a big network of interconnected objects, people, places, or organisations. A graph, for example, could represent Facebook as a giant web of people and relationships. Graphs can be huge. Advanced computing technology such as neural networking can help find patterns in this tangled mass of data. Product recommendation engines like those used by Netflix and typically rely on graph analysis.

Inline Bots

Inline bots are bots that can join your one-to-one or group conversation. Just @mention them within a chat for a seamless bot experience.

Intelligent agent

In artificial intelligence, an intelligent agent (IA) is an autonomous entity which observes through sensors and acts upon an environment using actuators (i.e. it is an agent) and directs its activity towards achieving goals (i.e. it is "rational", as defined in economics). Intelligent agents may also learn or use knowledge to achieve their goals. They may be very simple or very complex: a reflex machine such as a thermostat is an intelligent agent.
Intelligent agents are often described schematically as an abstract functional system similar to a computer program. For this reason, intelligent agents are sometimes called abstract intelligent agents (AIA) to distinguish them from their real-world implementations as computer systems, biological systems, or organizations. Some definitions of intelligent agents emphasize their autonomy, and so prefer the term autonomous intelligent agents. Still others (notably Russell & Norvig (2003)) considered goal-directed behavior as the essence of intelligence and so prefer a term borrowed from economics, "rational agent".

I.P.A. (Intelligent Personal Assistant)

An application that lets people ask questions by speaking in their natural language and listening to verbal answers. Although virtual assistants are on tablets and desktop computers, they caused the smartphone to become an incredibly useful electronic companion. No matter where people are, they can ask about anything that is public knowledge and gets an answer. Also called an “intelligent personal assistant,” “intelligent agent,” “digital assistant,” “personal assistant” and “voice assistant,” Apple’s Siri popularised the concept in 2011. Siri was followed by Google Now in Android and Chrome devices, Cortana in Windows, Amazon Echo (Alexa) and Facebook M. See Siri and Cortana. Virtual assistants cannot only answer questions, but they can also make phone calls, calendar appointments, and reminders and set alarms. The more information kept on the phone, the more personalised the results. Virtual assistants are the first embodiment of artificial intelligence used on a daily basis by hundreds of millions of people. The virtual assistant is driven by one or more semantic knowledge bases on the Internet as well as the data in the user’s device. Over time, the virtual assistant becomes more tailored to the individual.


Line (styled as LINE) is a freeware app for instant communications on electronic devices such as smartphones, tablet computers, and personal computers. Line users exchange texts, images, video and audio, and conduct free VoIP conversations and video conferences. Today the popular messaging service is operated by Line Corporation, the Japanese arm of Naver Corporation.

Machine learning

Machine learning is the process through which a computer learns with experience rather than additional programming. Let’s say you use a program to determine which customers receive which discount offers. If it’s a machine-learning program, it will make better recommendations as it gets more data about how customers respond. The system gets better at its task by seeing more data.

Machine perception

Machine perception is the capability of a computer system to interpret data in a manner that is similar to the way humans use their senses to relate to the world around them. The basic method that the computers take in and respond to their environment is through the attached hardware. Until recently input was limited to a keyboard or a mouse, but advances in technology, both in hardware and software, have allowed computers to take in sensory input in a way similar to humans. Machine perception allows the computer to use this sensory input, as well as conventional computational means of gathering information to gather information with greater accuracy and to present it in a way that is more comfortable for the user. These include computer vision, machine hearing, and machine touch.

Messaging Platform

A messaging platform is a unique tool that enables Internet users to exchange messages for the purpose of human communications.— for example, Facebook Messenger.

Microsoft Teams

Microsoft Teams provides the enterprise security and compliance features you expect from Office 365, including eDiscovery and legal hold for channels, chats, and files. Tooltip about the availability of the features. Available in 181 markets and 19 languages, Microsoft Teams encrypts data at all times, at rest and in-transit, and includes multi-factor authentication to enhance identity protection.

Model NLP

The simplest definition of a model is a mathematical representation of relationships in a data set. A slightly expanded definition: “a simplified, mathematically formalised way to approximate reality (i.e. what generates your data) and optionally to make predictions from this approximation.” Models get complicated. A simple model might be illustrated with a two-axis graph, but if your data is more complex, the predictive model will be more complex. When you speak to your smartphone, for example, it turns your speech into data and runs that data through a model to recognise it. That’s right, Siri uses a complex speech recognition model to determine to mean.


A multi-channel marketing strategy offers customers different touch points (shops, website, live chat, emails, mobile apps, etc.), without interconnecting them, while providing the same service.

Natural language generation (NLG)

NLG technologies transform data into written or spoken language that we understand. As experienced through services like Apple's Siri and seen in movies like Space Odyssey and Her, this technology is what most people think of first when they imagine interacting with an AI platform. However, NLG is also behind marketing content such as push notifications, emails, and even short-form articles that you may not be aware are artificially generated.

Natural language processing (NLP)

NLP technology lets computers understand human languages. It underlies language translation services such as Google Translate, speech recognition services such as Apple's Siri, and text recognition capabilities for social-media sentiment analysis like Sysomos. Chatbots (definition above) from H&M and Chase Bank show the potential of NLP in marketing and customer service.

Neural networks

Neural networks are computing systems that mimic the structure of the biological brain. Most modern AI products are built on neural networks to enable different forms of machine learning. Different neural network architectures have different strengths. We highlight a few common architectures below; you can find more technical explanations at the Asimov Institute.

Convolutional networks (CNN)

CNNs—not to be confused with cable news networks!—are often used for image recognition. They power Google Images "leaping cat" searches, Facebook faces recognition, and Snapchat faces swapping.

Deconvolutional Networks (DN)

DNS are reversed CNNs. They let you enter words, say "leaping cat," to generate an image of a leaping cat.

Recurrent neural networks (RNN)

RNNs are good at speech recognition and handwriting recognition. (This is not the same as a recursive neural network, also confusingly called an RNN. We'll post an update when we've clarified what these are best used for.) 

Notification Bot

Notification bots, sometimes called as 9 AM bots, are bots which send notifications (usually of new information) at a set time. A good example is a bot created by Harvard Business Review which sends you a gist of an article every day at 9 AM.

Nouvelle AI

Nouvelle artificial intelligence (AI) is an approach to artificial intelligence pioneered in the 1980s by Rodney Brooks, who was then part of MIT artificial intelligence laboratory. Nouvelle AI differs from classical AI by aiming to produce robots with intelligence levels similar to insects. Researchers believe that intelligence can emerge organically from simple behaviours as this intelligence interacted with the "real world," instead of using the constructed worlds which symbolic AIs typically needed to have programmed into them.


An omnichannel marketing strategy implements the notion of ubiquity: customers can use different channels at the same time. There is an even stronger coherence between channels since omnichannel enables users to experiment channels simultaneously. For an omnichannel customer experience, customer data needs to be kept and used to allow them to complete each stage of their journey without any efforts.

Predictive marketing

Predictive marketing refers to all techniques used to predict a customer’s future actions. This relies on data collected from past and current behaviours on a website. Predictive marketing can help anticipate the loss of an inactive customer and recommend personalised products on an eCommerce website.

Push Notifications

The delivery of information from a software application to a smartphone, tablet or computer, without a request from the client. The user is encouraged to click on one or more call to actions via these push notifications but is also given the possibility to opt-out. Here are some of the possibilities offered to you via these push notifications: answer an email, go on an app, view a web page, etc.

Reinforcement learning

Reinforcement learning is an area of machine learning inspired by behaviourist psychology, concerned with how software agents ought to take actions in an environment so as to maximise some notion of cumulative reward. The problem, due to its generality, is studied in many other disciplines, such as game theory, control theory, operations research, information theory, simulation-based optimisation, multi-agent systems, swarm intelligence, statistics, and genetic algorithms. In the operations research and control literature, the field where reinforcement learning methods are studied is called approximate dynamic programming. The problem has been studied in the theory of optimal control, though most studies are concerned with the existence of optimal solutions and their characterization, and not with the learning or approximation aspects. In economics and game theory, reinforcement learning may be used to explain how equilibrium may arise under bounded rationality.

Rich Interactions

Unlike regular text messaging applications, the best chatbots also offer rich interactions. A good example would be to display buttons to the user during a conversation. The buttons would be used to decide between two different branches of conversation (e.g. Check Account Balance or Make a Payment).


A built-in “intelligent assistant” that enables users of Apple iPhone 4S and later and newer iPad and iPod Touch devices to speak natural language voice commands in order to operate the mobile device and its apps.

Supervised vs. Unsupervised Learning

Machine learning can take two fundamental approaches.

Supervised learning is a way of teaching an algorithm how to do its job when you already have a set of data for which you know “the answer.” Classic example: To create a model that can recognise cat pictures via a supervised learning process, you would show the system millions of pictures already labeled “cat” or “not cat.” 
Unsupervised learning is how an algorithm or system analyses data that isn’t labelled with an answer, then identify patterns or correlations.


Skype is an application that provides video chat and voice call services. Users may exchange such digital documents as images, text, video and any other, and may transmit both text and video messages. Skype allows the creation of video conference calls.


Slack is a cloud-based team collaboration tool founded by Stewart Butterfield. Slack began as an internal tool used by their company, Tiny Speck, in the development of Glitch, a now defunct online game. The name is an acronym for "Searchable Log of All Conversation and Knowledge". Slack teams allow communities, groups, or teams to join through a specific URL or invitation sent by a team admin or owner. Although Slack was meant for organisational communication, it has been slowly turning into a community platform, a function for which users had previously used message boards or social media such as Facebook or LinkedIn groups. Many of these communities are categorised by topics which a group of people may be interested in discussing.


SnatchApp is a freeware, cross-platform, and end-to-end encrypted instant messaging and calling app available for Desktop, Windows 10, Android, Windows 8, iOS, I-Watch, and other smartphones. Just by a simple click, you can easily stay in contact with your family members, friends, and loved ones without any difficulty. Easy to download, Convenient To Use, Secure Conversations, Secure Group Chats, Highly Reliable, and much more are some of the unique features of this app. It not only offers fast and easy face-to-face conversations but also allows you to deliver expressive stickers anywhere across the country.


The goal of SnatchBot, since its inception in 2015, has been to expand the accessibility of chatbots and make bot-building easy for anyone, developer or otherwise. Based in the heart of the “start-up nation”, Herzliya Pituach, SnatchBot provides access around the world to sophisticated, natural-language conversational bots (“chatbots”) with highly engaging user experiences and lifelike conversational interactions across all communication channels.

SnatchBot Store

SnatchBot Store is a marketplace that is fully integrated with SnatchBot's comprehensive bot-building platform. It provides for free ready-for-use templates for a wide variety of chatbot use cases, including, but not limited to, customer service, banking, travel and tourism, and e-commerce.

SnatchBot Store is comprised of two parts: Bot templates and Bot directory. In the Bot template section, enterprises can access customizable, task-specific, and pre-built bots in a variety of areas and sectors. Users can simply choose a template that fits their needs and begins using their bot immediately. The bot directory includes established bots which can be added to your preferred channels with a single click.

Speech recognition

Speech recognition (SR) is the interdisciplinary sub-field of computational linguistics that develops methodologies and technologies that enable the recognition and translation of spoken language into text by computers. It is also known as "automatic speech recognition" (ASR), "computer speech recognition", or just "speech to text" (STT). It incorporates knowledge and research in the linguistics, computer science, and electrical engineering fields.


Telegram is a free cloud-based instant messaging service. Telegram clients exist for both mobile (Android, iOS, Windows Phone, Ubuntu Touch) and desktop systems (Windows, MacOS, Linux). Users can send messages and exchange photos, videos, stickers, audio, and files of any type. Telegram also provides optional end-to-end encrypted messaging.


ThingTon is a chatbot which acts as a conversational interface for the Internet of Things. It helps you interact with Smart Home devices via text messages.


Twilio (pronounced TWILL-e-o) is a cloud communications platform as a service (PaaS) company based in San Francisco, California. Twilio allows software developers to programmatically make and receive phone calls and send and receive text messages using its web service APIs. Twilio's services are accessed over HTTP and are billed based on usage.


Viber is a free, cross-platform instant messaging and voice over IP (VoIP) application that was first developed and popularised by the Israeli company Viber Media that was bought by the Japanese multinational company Rakuten. In addition to instant messaging, users can exchange images, video, and audio media messages by sending files to each other. As of December 2016, Viber had 800 million registered users.

Virtual Assistant

Virtual assistants are “smart” robots that use AI to help you complete tasks.


IBM Watson is a question-answering computer system which can answer questions which are posed in natural language. It became famous for winning Jeopardy competing against human contestants in 2011.

Web Bubbles (Bubbles)

Wubbles are the web within chat. They function the same way a web page would, but entirely within your one-to-one or group conversation.


WeChat (literally: "micro message") is a free, cross-platform and instant messaging application developed by Tencent. It was first released in January 2011 and was one of the largest standalone messaging apps by monthly active users. As of May 2016, WeChat has over a billion created accounts, 700 million active users; with more than 70 million outside of China (as of December 2015). In 2016, WeChat reached 864 million active users.

Next Steps

Learn everything you must require to use our API.