• the use of our admin console to easily create and manage all your content
• chatbox customisation, with our dydubox
• access to pre-defined analytical reports to monitor your virtual agent’s ongoing performance 
• a ready-to-use knowledge base for Expert chatbots

Our pricing mainly depends on the expected number of dialogues. 

Our subscriptions are on an annual, renewable basis and include:

• the licence to use the solution and its advanced natural language processing algorithm
• standard shared SaaS hosting in France (also available in dedicated version on AWS and Azure in the European Union) 
• access to all our online documentation and video tutorials
• access to our support service, on working days between 9:30am and 6pm
• regular project follow-up

Optional 

Depending on your needs, there are options to enable certain advanced features, such as:

• duplicating the use of your bot in another language or for different consultation spaces
• enabling bulk data import / export (always in compliance with GDPR)
• managing unlimited user profiles

During the implementation phase:

• training, knowledge base creation support, use of advanced features, connection to APIs, etc.
• continuous monitoring for the optimisation and improvement of your bot, our teams can offer personalised support

At dydu, the bot administrator creates the knowledge base from scratch, so it’s not really machine learning. Our solution also doesn’t learn alone, we use a supervised learning mode.

Supervised learning here means that the action of a human, in this case the bot’s administrator, is required to ensure that the bot understands your users. We use supervised learning modules. These modules aim to help the bot administrator manage and improve the business knowledge articles by providing suggestions for improvements according to matching, suggestions for knowledge that is too close, quality alerts, audit reports, etc.

Bot management by a team of operational staff, with a well-defined scope, allows to properly manage the answers provided to users and to obtain a high level of user satisfaction from the start, unlike machine learning that requires a lot of data, and therefore dialogues, so that the bot can learn by itself.

The distance calculation method means that an algorithm directly calculates the distance between the user’s questions and the sentences in the knowledge base. The closer the user’s question is to a reference (same words/lemmas, in the same order or not), the higher the matching score. Read our article to find out more about matching scores.

Supervised learning here means that the action of a human, in this case the bot’s administrator, is required to ensure that the bot understands your users. We use supervised learning modules. These modules aim to help the bot administrator manage and improve the business knowledge articles by providing suggestions for improvements according to matching, suggestions for knowledge that is too close, quality alerts, audit reports, etc.

Dydu also owns its own matcher, meaning that we designed our own algorithm to match two elements together. To put it simply, our linguistic component calculates the distance between the user’s request and all the sentences in your chatbot’s knowledge base. If the distance is low, the chatbot can provide an answer to the question. This enables the chatbot to provide far more relevant answers than those using a key words system. This operating mode also enables to quickly identify requests that need to be added to the knowledge base, without constantly having to check what already works.

Other external matchers can be used in addition for our SaaS customers, Microsoft Luis and Q&A.

Supervised learning here means that the action of a human, in this case the bot’s administrator, is required to ensure that the bot understands your users. We use supervised learning modules. These modules aim to help the bot administrator manage and improve the business knowledge articles by providing suggestions for improvements according to matching, suggestions for knowledge that is too close, quality alerts, audit reports, etc.

The more the chatbot is used, the more it can learn with features that aim to help the bot administrator manage and improve its knowledge articles.

Using an external matcher can increase the chances of direct matching, i.e., providing the user with a correct answer and therefore increasing customer satisfaction. However, this is less true for French-language bots, where dydu is the market leader thanks to a highly effective matcher. External matchers could be useful, in addition to dydu’s matcher, for other languages though, for which our NLP (Natural Language Processing) component is less well-trained. This could make your bot even more effective.

If you are interested in using external matchers, especially for languages other than French, we can offer Microsoft Luis and Q&A in addition to our dydu matcher.

When users ask a question and exchange with the bot, they can therefore answer in natural language or click on one of the buttons (bot suggestions). These are guided choices.

Let’s take the example of the chatbot on our webpage. If you ask, “what are your offers?”, the bot will suggest 3 buttons to guide the user.

In SaaS (Software as a Service) mode, we host your project on dedicated and shared servers. We guarantee a software update every 15 days.

However, if you opt to host on your own servers, in on-premise mode, we provide software LTS (Long Term Support) every three months. You do the upgrade yourself.

Ask for a free quote to find out everything about the cost of hosting your project in SaaS mode !

Our BMS is intuitive, effective and easy-to-use for everyone. You don’t need IT skills to make your bot a success, but business expertise and knowledge of your customers, employees and users’ needs, in order to provide them with the right answers and solutions.

During the bot implementation phase, our project managers will train you on how to use the BMS and provide you with the best practices for building a knowledge base.

Once your operational staff have been trained to use our user interface, they can create and manage the chatbot completely autonomously.

The time spent managing your bot depends on its scope and your own organisation. It is important to spend as long as necessary on bot management, especially at the beginning.

Regular bot management will also enable you to instantly adapt to difficult situations. 70% of our bot users (whether internal or customer bots) reported that their chatbots were even more useful during a crisis (Chatbot Observatory, April 2019 and March 2020).

Your CSM will :

• Help you improve your knowledge base (use of features, optimisation of your existing knowledge)
• Monitor your chatbot: KPI monitoring, extraction and formatting of analytics, preparation and holding of steering committees
• Carry out a scoping of desired / required changes
• Share their expertise

Depending on your needs, we can also offer :

• Bot performance and knowledge base audits
• Training courses
• Workshops
• Technical expertise: chatbox changes, additional API settings, server migration, etc.

This support is optional and comes at an extra cost.

Chatbot customisation includes :

• Text elements: your bot’s name, GDPR disclaimer, onboarding, header title, banner, welcome sentence, etc.
• Graphic elements: font, colours, avatar, logo, etc.
• Features: enlarge, reduce, collapse buttons, “copy conversation” and “contact” tabs, feedback, language display, character counter, etc.
• Your brand’s tone: language, jokes, emojis or gifs, etc.

You can deploy your chatbot autonomously and configure and customise the front-end of your bot with the dydubox interface. This enables you to easily try out different chatbox designs, text and features, and to change them whenever you want.

Additional answers

This enables you to complete an answer or to configure an action, based on data provided during the dialogue. The bot identifies pre-defined words and triggers an action.

For example, the bot could trigger an escalation to a human operator after 3 misunderstandings or if the user utters the words “terminate my contract”.

Alternative answers

The bot can answer the same question differently if it is asked several times. This improves the conversational experience, by making the bot sound more natural.

For example, if a user asks “how to contact customer services” twice in the same conversation, the chatbot can provide the same answer phrased differently, because it has alternative answers for this question.

Push rules (behavioural targeting)

This feature allows to observe the end-user’s browser behaviour and trigger an action at the right moment. For example, the chatbot could pop up and offer to help if the user is inactive on the site for more than X minutes.

Context conditions

This feature allows to condition a bot’s answer according to the value of a variable. To answer the question “I’m 21 years old, can I vote?”, the bot tests the value of the age variable (greater than or equal to 18 years) and provides the right answer according to this variable, instead of a generic answer, such as “any French citizen can vote from the age of 18”.

Consultation spaces

This enables the chatbot to adapt the answer to the user, depending on criteria such as which device they are using, or whether or not they are logged into their customer account (different answers to the same question according to these criteria).

Let’s take the example of a bot with two consultation spaces (mobile and PC) and a user who asks, “how to share my schedule”. If the user is asking this question from a mobile phone, the answer will be in text form. But if the user is using a PC, then the chatbot will provide a step-by-step tutorial in the sidebar.

Variables

User data can be retrieved during a conversation and used in an answer or API call. These variables can come from an explicit user declaration, a transfer to the chatbot (via its web page), or information from an API (e.g., name, age, type of contract, etc.).

Web services

By using web services from your chatbot’s knowledge base, you can access external information, such as tracking an order, or push information out to feed a CRM or ticket management tool, for example. If the bot is connected to a specific company app (CRM, IS, HRIS, etc.) via a web service, it can use the data to provide personalised answers.

]]> https://www.dydu.ai/en/questions/features/what-features-does-your-software-include/?utm_source=rss&utm_medium=rss&utm_campaign=what-features-does-your-software-include Thu, 06 Jan 2022 15:59:14 +0000 https://www.dydu.ai/?post_type=dydu_faq&p=746060 Our software is the fruit of a decade of research in natural language processing, to enable our clients to create high-performing bots that can hold a smooth conversation and provide generic or personalised answers according to the user and context.

Some features aim to improve the end users’ experience (front), whereas others aim to improve the bot manager’s experience via the Bot Management System (BMS) (back).

Features to improve the end-user’s experience

• Understanding: Politeness and small talk, phrases, monitoring of activity and improvements
• Personalisation: authentication, variables, context conditions, behavioural targeting, web services
• Rich / natural conversations: additional / alternative answers, social base, rich content, accessibility

Features to improve the bot manager’s experience

• Bot management: WYSIWYG*, knowledge article status, comments
• Chatbox customisation via the dydubox: design, informative texts (such as GDPR disclaimer, introduction, etc.), features, channels
• Improvement: dashboard, analytics, matching knowledge, misunderstood sentences, audits, quality alerts

*WYSIWYG (What You See Is What You Get): a visual text editor that enables you to preview the text before it is published.

The bot’s use cases are defined alongside your project manager during the chatbot implementation. They will help you identify the features you need, and also train you to use them.

If your project requires a lot of customisation and advanced use cases, then you can access our advanced features.

However, you can also create a simple chatbot, using our software’s basic features to meet your needs and objectives.

To find out more, read our series of articles on key features according to your use case:

• Chatbots and Recruitment: Benefits and Keys to Success
• Customer Service Chatbots : Benefits and Keys to Success
• Chatbots and After-Sales-Services : Client testimonial

Dydu can currently understand 10 languages:

–        French

–        English

–        Spanish

–        Italian

–        Romanian

–        Portuguese

–        German

–        Dutch

–        Polish

–        Finnish

We recommend creating your bot in one language, deploying it, analysing the dialogues and making improvements, before translating into another language. This will enable you to identify the most frequently called upon knowledge articles and translate them first.

To translate a chatbot from a source language, you need to export your knowledge base, translate the articles while maintaining the structure, and import the translation. It is important to take great care when editing the file.

Once your chatbot has been deployed in several languages, it will detect your user’s browser or application language and select the most appropriate language by default. The user can also manually select a language in the chatbot, via the flag button.

Rich content includes :

• Formatting of bot responses: bold, italics, underlined, bullet points, colour text
• Integrating hypertext links, emojis, videos, images, gifs and documents (via hypertext links only) in your answers
• Sidebars: adding content outside of the answer bubble to provide additional information (with text, images, videos, step-by-step tutorials, etc.)

These features are available in the Bot Management System (BMS) when you create your knowledge base and are very easy to use.

Chatbots are usually integrated on a web page or on your entire intranet or extranet site via a floating button, for example. You can also deploy your chatbot on Google chat (Hangout), Messenger, Microsoft Teams, or mobile and desktop apps.

To integrate your bot on an external channel, you need a connector. A connector enables users to communicate with dydu bots on external channels, such as Google chat (Hangout), Messenger, MS Teams, etc. A connector is developed to match dydu features with those of the targeted channel. For example, how will text formatted in the back office be displayed on a channel like Slack? How are links displayed ?

Callbots are, by very definition, accessible by telephone (landline or mobile), via a phone number.

Voicebots can be accessed via connected objects that use voice, such as connected speakers, and can also be deployed as “voice chatbots”, to make them accessible for people with disabilities. They can be accessed in the same place as the chatbot (computer, tablet or mobile).