
Explore how the Flows platform enables tailored conversational solutions with generative and deterministic AI. Leverage data analyst assistants, API assistants, and Rag assistants to deliver accurate, contextual responses.
Create and configure an end-to-end flow in the flow builder, integrating rag assistants for Genexus and Globant AI documents, with automatic english-to-french translation and dynamic variables.
Create a flow named tourist agent that uses a context-aware orchestrator to route user queries between tourist attractions information and weather in any city, using conditional components and go-to nodes.
Design flows using generative and logical components, configuring assistant and Rag assistant nodes, variables, and conditional branches, with go to navigation and dynamic scripting.
Explore integrations to create a flow by configuring api, email, and spreadsheet options. Define url, method, headers, recipients, spreadsheet details, and service account credentials to manage responses and data flow.
Master building end-user flows with response components such as message, image, link, card, video, quick reply, date picker, dropdown, and carousel, enabling dynamic content, variables, and event-driven navigation.
Navigate the GEAI flow builder to manage flows, switch languages, monitor with real-time logs, test the flow via the demo page and interactive chat.
Learn to manage variables in flows by creating, naming, and configuring them in the flow builder’s variables option, using constant and protected options, types, and default values.
Explore the conversation history in the flow builder's logs diagnostics to search, filter by time, export to CSV, and view detailed user conversations with timestamps and full message logs.
Configure the flow builder by managing authentication keys, integrations with Slack and WhatsApp, and flow versions, with JSON import/export and restore options in the danger zone.
build a flow called internet search flow that uses two chat assistants: one relies on internet search results, the other rephrases questions from history, linked via a search API.
Expose a flow on a web page by generating an application key and embedding the JavaScript code to load the web SDK, set language, and render a customizable chat.
Design a mall customer assistant flow by integrating an orchestrator with a movie ticket sales assistant, mall general information assistant, and a fallback assistant to resolve queries and bookings.
Presentation
Flows is the proposal from Globant Enterprise AI to create conversational solutions tailored to the specific needs of each company, leveraging the capabilities of generative and deterministic artificial intelligence.
User interaction can be significantly improved, as it is not only possible to develop chatbots that respond accurately and appropriately, but it also facilitates the integration of specialized assistants, such as Data Analyst Assistants, API Assistants, Chat with data Assistants, and RAG Assistants.
Objective
Through different examples, we will teach you the necessary concepts to create and test flows using various interaction components. As a case study, we propose the creation of an agent that helps end users with inquiries related to the management of a Shopping Mall (inquiries about stores, services, movie listings, and ticket reservations).
Prerequisites
Although it is not a mandatory requirement, we suggest completing the "Assistants" module within the Globant Enterprise AI course beforehand.
Suggested Methodology
Watch each video and try to replicate the examples developed there. For this, you will need an authorized user account to access Globant Enterprise AI.
To request it, you can go here.
Approximate duration: 6 hours (watching the course videos and actually completing the course) Watching the videos alone takes 1 hour 40 minutes