Block libraries - Ready-made skills – Blip | Blip Help

Introduction

The block library was created to facilitate the day-to-day of those who need to build an intelligent contact without having deep technical mastery and need speed to publish. With it, it will be possible to apply both basic deterministic skills such as Onboarding, LGPD, Inactivity, Validation cascade, Something else, Finalization, and CSAT as well as structures using builder agents.

Basic concepts

The structure for using ready-made blocks is designed in the router architecture for proper functioning, where each set of ready-made blocks will stay within a builder/Studio connected to a router. In this way, the intelligent contact is easy to maintain and evolve.

Is it possible to use ready-made blocks within a single builder/Studio? Yes, it is possible, but there will be a challenge to perform evolutions and maintenance due to the interconnection between the blocks, where if you need to edit a specific part of the structure, it will be necessary to understand which other parts are connected to the part that will be edited or evolved, thus causing a high expenditure of time to configure.

Macro view

Before starting to use the library, it is important to have the conversational structure that will be made well-defined and how the parts of this conversation will interact with each other. Normally, a conversation has a beginning, middle, and end, following a main objective. The macro view is very similar to BPMN flowcharts. In this view, it is important to visualize the paths, the decision triggers, and to have symbols that represent the events within the conversation structure.

The library

The block library is divided into Skills, Agents, and Use Cases.

Skill is a set of standard blocks with a specific objective within the intelligent contact conversation.
Agent can be an agent block or a set of agent blocks and standard blocks, where agent blocks are fundamental for the specific objective within the intelligent contact.
Use case is a set of skills and agents for a specific objective within an intelligent contact journey.

Each option within the library has a tag and can be searched in the search bar as represented in the image below:

Ready-made blocks

Skills:
Onboarding LGPD Inactivity Opt-In/Out Something Else	Finalization CSAT Human Handoff Personal Data Validation Cascade

Agent:	Use case:
Registration	Lead Qualification

Skills

Onboarding

Guides and welcomes the user, explaining how to start using the service and performs greeting or return procedures and checks if the LGPD has already been answered.

LGPD:

Explains briefly what the LGPD is and collects the user's acceptance so that the Intelligent Contact can collect sensitive data.

Inactivity:

Sends a message to rescue the user's conversation with the Intelligent Contact after an inactive period.

Opt-In/Out:

Manages the user's authorization to receive or not receive communications.

Something Else:

Checks if the person wants to add information or talk about another subject.

Finalization:

Concludes the service in a clear way, confirming the closure of the interaction, being able to leave open the possibility of returning to the conversation.

CSAT:

Collects the user satisfaction evaluation after the service.

Human Handoff:

Forwards the service to a human when necessary.

Personal data:

Requests, validates, and organizes user's personal information securely. CPF - Name - Email - Phone

Validation Cascade:

Step-by-step process to confirm provided data, validating each piece of information before proceeding to the next.

Hierarchy of Validation
Attributes	Description
customized_errors	Used for customized validations that the other stages do not cover, e.g.: format of data requested by the bot.
api_error	Used to validate if the user was redirected from an HTTP call failure. If so, it should not proceed with the remaining stages of the cascade.
validation_media_error	Used to identify media errors.
audio_error	Adds a specific stage for audio validation, if it exists. E.g.: Blip speech to text.
link_error	Validates if the text contains only a link. When only text or text with links exists, it proceeds to the next stage. If it has only a link, it presents a message and returns to the previous skill.
emoji_error	Validates if the text contains only emojis. When only text or text with emojis exists, it proceeds to the next stage. If it has only emojis, it presents a message and returns to the previous skill.
botxbot_error	Validates if the user is a bot. If the last inputs (adjustable quantity) are identical OR the last input matches with the regex of automatic messages, the user is considered a bot. To proceed, they must send specific text; otherwise, the session is terminated.
non_talkables_error	Handling for sensitive terms. The variables must be defined: {{config.utilsApiUrl}}, {{config.utilsApiAuth}}, {{config.globalSheetsId}}, {{config.connectionTalksSheetsTabName}}, {{config.connectionTalksSheetsTabRange}}.
clear_text	Used for text cleaning before the regex stage is called to the NLP provider.
global_regex_error	Validação de regex global para inputs em qualquer ponto do bot.
long_message_error	Global regex validation for inputs at any point of the bot. Used to validate if the message is too long to be sent to NLP. Maximum character count currently: 100.
connection_talks_error	Functionality connected to the Smalltalks service. It is responsible for identifying a small talk and defining the message that will be sent to the user in case of a match. In addition, it is also possible to define the bot and block for the destination after receiving the message. The variables must be defined: {{config.utilsApiUrl}}, {{config.utilsApiAuth}}, {{config.globalSheetsId}}, {{config.nonTalkableSheetsTabName}}, {{config.nonTalkableSheetsTabRange}}.
short_message	Used to prevent calls to the NLP provider if the message is too short. Checks if it contains 3 or more letters.
nlp_error	Used if there is integration with an NLP provider in the bot. The configuration variables must be defined: {{config.AuthorizationApiKeyIA}} and {{config.commandsUrl}}.
content_assistent	Used if there is integration with a content assistant. The configuration variables must be defined: {{config.AuthorizationApiKeyIA}} and {{config.commandsUrl}}.

In the non_talkables_error and connection_talks_error attributes, there is a connection with external systems to search for a response according to the context and point of the conversation.

As soon as the cascade is added within the Studio area through the library, look for the block [REQ] set variables:

In the actions tab inside the Process ExceptionHierarchy script, it is possible to determine which validation will be active (true) or deactivated (false) within the conversation.

The non_talkables_error and connection_talks_error attributes are deactivated by default. If you want to activate them, just set to true and save. But remember to configure these validations according to your context.

Skills Configuration

Step by step:

Create a chatbot

Access the library

Adding a skill

Skills Instructions

Each skill imported into the builder area will have a block called "comece aqui" (start here). In this block, there will be configuration instructions for the imported skill; just apply the configurations and the skill will be working.

Normally, the welcome and standard error blocks, which already come as soon as the chatbot is created, can be deleted

Registration Agent

The ready-to-use agent within the block library is an agent for the registration of basic information such as name, e-mail, and phone; after the collection, it is directed to a standard block to continue the conversation.

Use case - Lead Qualification

Being a use case, a set of skills will all stay within a single builder as in the image below:

Using a single builder to place all the skills can facilitate due to the fact that everything is within a single work environment, but on the other hand, there will be a challenge if you need to evolve or add new things in the conversational structure: it will be to have redoubled care in the connections between the blocks.

To know more details on how to use and some tips for using the block library, we have an AMA on our YouTube channel.