Real-time Transmission Conversation (SSE)
Artificial Intelligence Chat (Real Time)The user receives the answers of artificial intelligence letter to letter, progressively and fluidly, without having to wait until the complete answer is ready for...
Informed Conversations
Help teams to get answers from the knowledge of the organization.
Agents and tools
Accompany tasks with wizards connected to their sources and tools.
More Ways to Talk
Take access to knowledge to the channels and languages used by your teams.
Key capabilities
A first look at what you can do with this product.
Document-based responses
Refer to documented knowledge through responses based on available sources.
Agents and tools
Accompany tasks with attendees who can consult information and use connected tools.
Custom Widgets
Adapt each conversation experience to the context and needs of your users.
Conversations with context
Keep the thread of the conversation to give continuity to each query.
WhatsApp, voice and languages
Attend interactions in more channels, by voice and in different languages.
Administration and control
Control the access, use and activity of attendees.
Explore our capabilities
Select a category in the map to explore its features.
features of IFINDIT CHATGEN
Choose a category or search across all product features. Browse the results page by page.
103 features
The search includes all categories.
All features
Artificial Intelligence Chat (Real Time) 10 features
Generation Progress Indicator
Artificial Intelligence Chat (Real Time)While AI prepares its response, the system displays progress messages that inform the user at what stage the process is at.
Automatic Storage of the History of Conversation
Artificial Intelligence Chat (Real Time)Each user message and each AI response are automatically saved in the conversation history to maintain context.
Chat Mode without Transmission (No Streaming)
Artificial Intelligence Chat (Real Time)In cases where real-time transmission is not adequate, the system can deliver the full response at one time.
Real Time Questions Suggestions
Artificial Intelligence Chat (Real Time)The system can automatically generate suggestions for questions related to the topic of the conversation, helping the user to continue the dialogue.
Intelligent Search Box (OneBox)
Artificial Intelligence Chat (Real Time)Special query mode that combines semantic search and generative response to offer direct results to specific questions.
Suggestions in the OneBox
Artificial Intelligence Chat (Real Time)The smart search box can also generate suggestions for related queries in real time.
Active Conversation Validation
Artificial Intelligence Chat (Real Time)The system verifies that the conversation exists and is active before processing any message, avoiding errors due to invalid or closed conversations.
Widget-Configuration Compatibility Validation
Artificial Intelligence Chat (Real Time)The system automatically checks that the widget type and AI settings are supported before starting any conversation.
Validation of Origin of Applications
Artificial Intelligence Chat (Real Time)The system verifies that applications come from authorized origins, blocking accesses from unauthorised sites.
Artificial Intelligence Agents 8 features
Simple Agent
Artificial Intelligence AgentsBasic type of agent that processes the user query and generates a direct response using the configured language model.
Agent React (Raisoning and Action)
Artificial Intelligence AgentsIt combines the processing of queries with the use of tools using a ReAct agent.
Planning Officer
Artificial Intelligence AgentsAgent who first elaborates a plan of steps to resolve the query and then executes them in order, ideal for complex multi-stage tasks.
Reasoning Agent
Artificial Intelligence AgentsIt includes a type of consultation-oriented agent that requires pre-response reasoning.
Asynchronous Execution of Agent Tasks
Artificial Intelligence AgentsThe complex tasks of the agents are processed in the background without blocking the user interface, which receives a notification when they finish.
Sending Tasks to Process Cola (Kafka)
Artificial Intelligence AgentsSend agents' tasks to a Kafka tail for asynchronous processing.
Task of Agent: Creation and Monitoring
Artificial Intelligence AgentsEach task assigned to an agent is recorded with a unique identifier that allows monitoring of its status (pending, in process, completed).
Agent Configured by Widget
Artificial Intelligence AgentsThe agent's type and behavior is automatically determined according to the settings of the chat widget, without requiring manual adjustments per session.
Tools of Agents (Tools) 6 features
Knowledge Base Tool
Tools of Agents (Tools)Agents can automatically consult the knowledge base to find relevant information to support their responses.
Mathematical Assessment Tool
Tools of Agents (Tools)Allows agents to use a mathematical calculation tool.
MCP Tool (Model Context Protocol)
Tools of Agents (Tools)Agents can connect to external MCP servers to access additional functionalities and data in a standardized manner.
Tool Call Record
Tools of Agents (Tools)Each time an agent uses a tool, the call is recorded with its parameters and result for audit and traceability.
Extendable Tool Framework
Tools of Agents (Tools)The system is designed to add new tools without modifying the main code of agents, using a common base interface.
Management of Tools from the Management Panel
Tools of Agents (Tools)The administrator can create, edit and delete tools available to agents from the management interface.
Integration with MCP Protocol 3 features
MCP Client
Integration with MCP ProtocolThe system can be connected as a client to external MCP servers to request additional capabilities in a standardized manner.
HTTP-STDIO bridge for MCP
Integration with MCP ProtocolThe system can communicate with MCP servers operating through standard input/output, using an HTTP bridge as an intermediary.
MCP Server Management
Integration with MCP ProtocolThe administrator can register, edit and delete the MCP servers that agents can use.
Generation Augmented Recovery (RAG) 5 features
Semantic Search on Knowledge Base
Generation Augmented Recovery (RAG)The system searches for relevant documents not by exact words, but by the meaning of the query, finding related information even if it does not use the same words.
Results Recomputer (Reranker)
Generation Augmented Recovery (RAG)After the semantic search, the system applies a second relevance filter that reorders the documents found to prioritize the most useful ones for the response.
Document-Based Response Generation (RAG)
Generation Augmented Recovery (RAG)It generates responses from documents recovered from a knowledge base using RAG.
RAG configuration by Widget
Generation Augmented Recovery (RAG)Each chat widget can have its own RAG settings, pointing to different knowledge bases depending on the case of use.
RAG Configuration Management
Generation Augmented Recovery (RAG)The administrator can create, edit and remove RAG settings from the management panel.
Knowledge Base and Vector Storage 5 features
Creating a Knowledge Base
Knowledge Base and Vector StorageThe administrator can create document repositories that agents and AI will use as a source of information to respond.
Vectorial Collections Management
Knowledge Base and Vector StorageThe documents of the knowledge base are organized into special collections that allow searches by meaning in an efficient manner.
Vector Storage Provider: ChromaDB
Knowledge Base and Vector StorageThe system uses ChromaDB as a vector storage engine, which allows searching for documents by semantic similarity.
File Storage in MinIO
Knowledge Base and Vector StorageStores knowledge base files in MinIO, compatible with S3 object storage.
Management of Vectorial Warehouses
Knowledge Base and Vector StorageThe administrator can create, configure and remove the available vector stores for knowledge bases.
Language Models (LLM) 6 features
Language Model Management
Language Models (LLM)The administrator can register and configure different language models (LLM) that chatbots can use.
Selection of LLM by Chat Settings
Language Models (LLM)Each chatbot can be configured to use a specific language model, allowing you to optimize costs and capabilities.
Tokens Use Monitoring
Language Models (LLM)The system records how many tokens (text units) consume each conversation, allowing control of LLM usage costs.
LLM Response Cache
Language Models (LLM)The system can temporarily store frequent responses from the language model to deliver them faster without generating them again.
Cache Configuration by LLM
Language Models (LLM)The administrator can configure whether and how the answers of each language model are cached.
Response Cache with Redis
Language Models (LLM)Use Redis to store and retrieve cached responses.
Prompts System and Response Templates 5 features
System Prompts Management
Prompts System and Response TemplatesThe administrator can create and edit the basic instructions that guide AI behavior in each chatbot.
Prompting Settings Management
Prompts System and Response TemplatesIt allows you to configure how system prompts are built and combined with user messages.
Response Templates
Prompts System and Response TemplatesThe administrator can define predefined response formats that AI can use to structure their responses consistently.
Sanitization of the Assistant's History
Prompts System and Response TemplatesThe system automatically cleans the history of AI before sending it as context, eliminating technical labels or noise that could confuse the model.
Automatic Improvement of Consultations (Enhancement)
Prompts System and Response TemplatesThe system can automatically improve the user query before processing it, making it clearer and more complete for the model.
Chat and Widget Settings 10 features
Creating Chat Settings
Chat and Widget SettingsThe administrator can create settings that define the complete behavior of a chatbot: type of AI, model, prompts, tools and more.
Types of Application IA: RAG and Agent
Chat and Widget SettingsThe system supports two main chatbot modes: RAG (responds based on documents) and Agent (use tools and reasoning to solve tasks).
Creating and Managing Widgets
Chat and Widget SettingsIt allows you to create and manage chat widgets for insertion into websites and applications.
Chat Type Widget
Chat and Widget SettingsInteractive conversation Widget that allows users to chat with AI naturally.
Widget Test from Panel
Chat and Widget SettingsThe administrator can test the operation of a widget directly from the management panel before publishing it.
Control of Origins Allowed by Widget
Chat and Widget SettingsThe administrator can define from which websites each widget can be used, blocking unauthorized accesses.
Widget-Configuration Compatibility
Chat and Widget SettingsThe system has compatibility rules that determine which AI settings are valid for each widget type.
Initialization of Compatibility Rules
Chat and Widget SettingsWhen starting, the system automatically loads the default compatibility rules if they do not already exist.
Widget Initial Message (Welcome Message)
Chat and Widget SettingsEach widget can be configured with a welcome message displayed to the user at the start of the conversation.
Input Interface Management
Chat and Widget SettingsThe administrator can configure which types of input the widget accepts: text, voice, files, etc.
Conversation Management 6 features
Creation of Conversations
Conversation ManagementUsers can start new chatbot conversations from the widget or management interface.
History of Conversation
Conversation ManagementThe system saves the complete history of each conversation, allowing you to return to the thread of dialogue in future sessions.
Conversation Context Management
Conversation ManagementThe system maintains and manages the context of the conversation so that AI remembers what has been said above.
Automatic Language Detection
Conversation ManagementThe system can automatically detect the language in which the user writes and adapt the response to the same language.
Anonymous Users in Conversations
Conversation ManagementSupports non-Login conversations and manages temporary identifiers for those users.
Configurable Conversation Context
Conversation ManagementThe administrator can configure how many previous messages are included as context for each new AI response.
Content Filtering 3 features
Inappropriate Content Filter (Profanity)
Content FilteringThe system can automatically detect and filter messages with inappropriate language or offensive content before processing them.
Filter List Management
Content FilteringThe administrator can configure the words and patterns that will activate the content filter.
Custom Response to Filtered Content
Content FilteringWhen a message is filtered, the system can respond with a custom message instead of ignoring the request.
Related Data Infrastructure (RDB) 4 features
Related Database Management
Related Data Infrastructure (RDB)The administrator can record relational databases (such as PostgreSQL) that agents can consult for structured information.
Database Table Management
Related Data Infrastructure (RDB)It allows you to configure which database tables are accessible to agents.
Table Column Management
Related Data Infrastructure (RDB)It allows to define which columns of each table are visible and queryable by the agents.
Related Data Referral by Agents
Related Data Infrastructure (RDB)Agents can consult relational databases for structured information to support their responses.
Transcription and Translation 2 features
Transcription of Audio to Text
Transcription and TranslationThe system can convert voice or audio messages into text to process them as queries to the chatbot.
Automatic Message Translation
Transcription and TranslationThe system can automatically translate user messages or AI responses into different languages.
Integration with WhatsApp 4 features
Message Producer for WhatsApp
Integration with WhatsAppThe system has a dedicated service that receives messages from WhatsApp and sends them to the processing queue of the chatbot.
WhatsApp Webhook
Integration with WhatsAppThe system exposes a connection point (webhook) that WhatsApp uses to send users' messages.
WhatsApp Incoming Message Processing
Integration with WhatsAppMessages received by WhatsApp are automatically processed and answered by artificial intelligence.
Answer Automatic via WhatsApp
Integration with WhatsAppAfter processing the user message, the system sends the response back to the user via WhatsApp.
API and Security Keys 5 features
API Key Management
API and Security KeysThe administrator can create and manage access keys that allow external applications to connect to the chat service.
Multi-tenant authentication (Multi-tenant)
API and Security KeysIt supports several organizations with data separation and tenant configuration.
Tenants (Tenants)
API and Security KeysThe administrator can create and manage organizations that use the chat platform.
Sensitive Data Encryption
API and Security KeysSensitive credentials and data stored in the database are automatically encrypted to protect them.
Public routes without Authentication
API and Security KeysThe system distinguishes between routes requiring authentication and accessible public routes without credentials, such as the endpoint of the chat widget.
Asynchronous Processing and Task Cola 6 features
Task Cola with Apache Kafka
Asynchronous Processing and Task ColaSend intensive processing tasks to Apache Kafka to run in the background.
Consumer of Asynchronous Tasks
Asynchronous Processing and Task ColaAn independent service continuously reads the tail tasks and executes them in the background.
Fallen Message Tail (DLQ)
Asynchronous Processing and Task ColaDirects tasks that fail to a queue of failed messages for review.
Consumer of Database Events
Asynchronous Processing and Task ColaA dedicated service processes the change events in the database to keep the different system components synchronized.
Database Event Processor
Asynchronous Processing and Task ColaTransforms and applies the events received from the database, running the necessary operations in response to changes.
Consumer Dynamic Reloading
Asynchronous Processing and Task ColaThe consumer service can be recharged hot without needing to restart the entire system when the configuration changes.
Task Programmer (Scheduler) 3 features
Task Programming Service
Task Programmer (Scheduler)The system includes a planner that executes automatic tasks in predefined times, without manual intervention.
Configurable Periodicals
Task Programmer (Scheduler)The administrator can define which tasks are executed automatically and how often.
Autonomous Implementation in Second Plane
Task Programmer (Scheduler)The planner operates independently to the main server.
Management Panel (Frontend TypeScript) 8 features
Chat Configuration Module
Management Panel (Frontend TypeScript)Interface to create, edit and manage chatbot settings, including model, prompts and tools.
Widget Management Module
Management Panel (Frontend TypeScript)Interface to create, configure, test and publish chat widgets that are inserted into websites.
Knowledge Base Module
Management Panel (Frontend TypeScript)Interface to manage the databases of documents that feed the chatbot responses.
Data Infrastructure Module
Management Panel (Frontend TypeScript)Interface to record and manage relational databases and vector warehouses connected to the system.
Conversation Module
Management Panel (Frontend TypeScript)Interface to view and manage the conversation history of all users.
Audit module
Management Panel (Frontend TypeScript)Interface to review activity records, errors and system usage.
LLM Module (Language Models)
Management Panel (Frontend TypeScript)Interface to record and configure available language models for chatbots.
Widget Test Page
Management Panel (Frontend TypeScript)Dedicated page on the panel where the administrator can test the chat widget in real time before posting it.
Architecture and Operations 4 features
Multi-Dominion Architecture
Architecture and OperationsThe backend is organized into independent functional domains: LLM engine, chat manager, widgets and identity, facilitating the development and independent maintenance of each...
Multi-Rent API
Architecture and OperationsThe API supports multiple organizations and separates their data and settings per tenant.
Production Server with Gunicorn
Architecture and OperationsIn production, the system uses Gunicorn as a high-capacity web server, with customizable configuration of workers.
Deployment with Docker and Docker Compose
Architecture and OperationsIt allows you to deploy services through Docker and Docker Compose.
We didn't find features with those terms. Try a different word.
Part of the IFINDIT platform
Explore how it relates to other products and shared capabilities.
Persons and equipment
Consultation · Analysis · Information management
Sources of information
Documents and Repositories · Databases · Applications and APIs
A functional view of IFINDIT and its shared capabilities. Each product connects different needs of the organization.
View platform architecture (SVG)
Let's talk about your information.
Tell us what you need to find, analyze or organize. Discover IFINDIT with the Creangel team.