Thousands of companies use Botpress to build amazing chatbots.
We're proud to be considered the de-facto standard platform for conversational AI by our community of developers.
We benchmark and fine-tune the latest NLP models specifically optimized for domain-specific, contextual and goal-oriented conversations.
Our Language Understanding engine is not only the most advanced, but also the easiest to use for developers.
import axios from 'axios'import trainData from './train_data.json'const hostname = "http://localhost:3200"const trainUrl = hostname + "/v1/train"const { data } = await axios.post(trainUrl, { ...trainData, password: "123456" })console.log(data) // Prints { modelId: "fk2j-9394-0kfj-g385-76dj" }
import axios from 'axios'const hostname = "http://localhost:3200"const modelId = "<PUT MODEL ID HERE>"const predictUrl = hostname + "/predict/" + modelIdconst { data } = await axios.post(predictUrl, { password: "123456", utterances: ["Fly to Dubia please"]})// data.spellChecked: "Fly to Dubai please"// data.intents: [{ name: 'book_flight', confidence: 0.98 }, ...] // data.slots: { destination: { canonical: "DXB", source: "Dubai" } }// data.entities: [ { entity: "airport", canonical: "Dubai" } ]
{ "language": "en", "intents": [ { "name": "book_flight", "contexts": [ "global" ], "utterances": [ "book flight", "book flight to [quebec](destination)", "let's go to [new york](destination) tomorrow", ], "slots": [{ "name": "destination", "entities": [ "airport" ] }] }, { "name": "hello", "contexts": [ "global" ], "slots": [], "utterances": [ "good day!", "good morning", "holla", "bonjour", "hey!", "hi", "hello" ] } ], "contexts": [ "global" ], "entities": [ { "name": "airport", "type": "list", "fuzzy": 0.9, "values": [ { "name": "JFK", "synonyms": [ "New York City", "New York" ] } ] } ]}
Botpress exposes a simple API for training, and one for prediction. No need for an ML expert, pipeline configuration, or hyper-parameters.
Automatic spelling corrections to better understand users.
Accurately determine the user’s intent from their utterance.
Automatically identify the language being used.
Like Airport code from a city name, or Date from ‘tomorrow morning’ - To create actional information from the user utterances.
Automatically gather key information to complete the user’s request.
We actively maintain integrations with the most popular messaging services including Facebook Messenger, Slack, Microsoft Teams and Telegram.
Looking to embed a chat window on your website or in your mobile application? Botpress Webchat is a feature-rich chat that can be embedded anywhere.
Building a conversational experience that customers will love is hard, even when equipped with the best NLU technology. That’s why we created the Conversation Studio: a powerful visual tool to build “awesome-by-default” conversations.
Building a chatbot with Botpress feels no different than building any other type of application.
Developers can extend Botpress by building custom integrations using our powerful Botpress SDK.
Botpress Enterprise provides out-of-the-box integrations with many enterprise software and industry-standard security protocols such as SAML and LDAP.
How businesses use Botpress to build better chatbots
“Botpress is the centerpiece of ASISPO, helping us to understand patients’ questions and to give them the right contextualized and personalized answer.”
-Thomas Gouritin, Co-Founder, ASISPO
“Botpress provides a great backbone of the software solution that we provide for our customers.”
-David Mika, Innovation and Digital Transformation, VR Bank
“The Botpress platform is extremely easy to use and has all of the necessary functionalities that we require to bring automations to our customers”
-Valerio Lombardi, CEO, engineon