ActionKit
ActionKit
One API to equip your AI
One API to equip your AI
agent
agent
with 5,000+ integration tools
with 5,000+ integration tools
Trusted by leading AI SaaS engineering teams
Trusted by leading AI SaaS engineering teams
Trusted by leading AI SaaS engineering teams
130+
130+
integration connectors
integration connectors
5000+
5000+
3rd-party tools
3rd-party tools
>10M
>10M
executions per day
executions per day
How it works
How it works
Authenticate your users and give your AI agent product thousands of 3rd-party tools
Authenticate your users and give your AI agent product thousands of 3rd-party tools
Your customers connect their external apps in your product
Your customers connect their external apps in your product
Paragon’s embedded UI components provide a fully managed authentication flow for your users, directly in your product.
Paragon’s embedded UI components provide a fully managed authentication flow for your users, directly in your product.
Add ActionKit as a function tool
Add ActionKit as a function tool
With just a few lines of code, you can add ActionKit as an available tool for your AI agent.
ActionKit is compatible with any LLM framework.
With just a few lines of code, you can add ActionKit as an available tool for your AI agent.
ActionKit is compatible with any LLM framework.
// Langchain import { ChatOpenAI } from '@langchain/openai' import { createToolCallingAgent } from 'langchain/agents' import { createLangChainTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); createToolCallingAgent({ llm: new ChatOpenAI({ model: 'gpt-4o-mini', temperature: 0 }), tools: createLangChainTools(actions); prompt: ChatPromptTemplate.fromMessages([ ['placeholder', '{chat_history}'], ['human', '{input}'], ['placeholder', '{agent_scratchpad}'] ]) });
// OpenAI SDK import requests from openai import OpenAI client = OpenAI() actions_url = f"https://actions.useparagon.com/projects/{project_id}/actions" get_actions_params = { "categories": "crm,project_management" } response = requests.get(actions_url, params=params, headers=actions_auth_header) paragon_tools = response.json() messages = [{"role": "user", "content": "Help me create a Jira ticket"}] completion = client.chat.completions.create( model="gpt-4o", messages=messages, tools=paragon_tools, tool_choice="auto" )
// Vercel AI SDK import { generateText } from 'ai'; import { openai } from '@ai-sdk/openai'; import { createAISDKTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); // Tool calling from available tools await generateText({ model: openai('gpt-4o'), tools: createAISDKTools(actions), toolChoice: 'auto', prompt: 'Help me create a new task in Jira.' });
// Langchain import { ChatOpenAI } from '@langchain/openai' import { createToolCallingAgent } from 'langchain/agents' import { createLangChainTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); createToolCallingAgent({ llm: new ChatOpenAI({ model: 'gpt-4o-mini', temperature: 0 }), tools: createLangChainTools(actions); prompt: ChatPromptTemplate.fromMessages([ ['placeholder', '{chat_history}'], ['human', '{input}'], ['placeholder', '{agent_scratchpad}'] ]) });
// OpenAI SDK import requests from openai import OpenAI client = OpenAI() actions_url = f"https://actions.useparagon.com/projects/{project_id}/actions" get_actions_params = { "categories": "crm,project_management" } response = requests.get(actions_url, params=params, headers=actions_auth_header) paragon_tools = response.json() messages = [{"role": "user", "content": "Help me create a Jira ticket"}] completion = client.chat.completions.create( model="gpt-4o", messages=messages, tools=paragon_tools, tool_choice="auto" )
// Vercel AI SDK import { generateText } from 'ai'; import { openai } from '@ai-sdk/openai'; import { createAISDKTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); // Tool calling from available tools await generateText({ model: openai('gpt-4o'), tools: createAISDKTools(actions), toolChoice: 'auto', prompt: 'Help me create a new task in Jira.' });
// Langchain import { ChatOpenAI } from '@langchain/openai' import { createToolCallingAgent } from 'langchain/agents' import { createLangChainTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); createToolCallingAgent({ llm: new ChatOpenAI({ model: 'gpt-4o-mini', temperature: 0 }), tools: createLangChainTools(actions); prompt: ChatPromptTemplate.fromMessages([ ['placeholder', '{chat_history}'], ['human', '{input}'], ['placeholder', '{agent_scratchpad}'] ]) });
// OpenAI SDK import requests from openai import OpenAI client = OpenAI() actions_url = f"https://actions.useparagon.com/projects/{project_id}/actions" get_actions_params = { "categories": "crm,project_management" } response = requests.get(actions_url, params=params, headers=actions_auth_header) paragon_tools = response.json() messages = [{"role": "user", "content": "Help me create a Jira ticket"}] completion = client.chat.completions.create( model="gpt-4o", messages=messages, tools=paragon_tools, tool_choice="auto" )
// Vercel AI SDK import { generateText } from 'ai'; import { openai } from '@ai-sdk/openai'; import { createAISDKTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); // Tool calling from available tools await generateText({ model: openai('gpt-4o'), tools: createAISDKTools(actions), toolChoice: 'auto', prompt: 'Help me create a new task in Jira.' });
// Langchain import { ChatOpenAI } from '@langchain/openai' import { createToolCallingAgent } from 'langchain/agents' import { createLangChainTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); createToolCallingAgent({ llm: new ChatOpenAI({ model: 'gpt-4o-mini', temperature: 0 }), tools: createLangChainTools(actions); prompt: ChatPromptTemplate.fromMessages([ ['placeholder', '{chat_history}'], ['human', '{input}'], ['placeholder', '{agent_scratchpad}'] ]) });
// OpenAI SDK import requests from openai import OpenAI client = OpenAI() actions_url = f"https://actions.useparagon.com/projects/{project_id}/actions" get_actions_params = { "categories": "crm,project_management" } response = requests.get(actions_url, params=params, headers=actions_auth_header) paragon_tools = response.json() messages = [{"role": "user", "content": "Help me create a Jira ticket"}] completion = client.chat.completions.create( model="gpt-4o", messages=messages, tools=paragon_tools, tool_choice="auto" )
// Vercel AI SDK import { generateText } from 'ai'; import { openai } from '@ai-sdk/openai'; import { createAISDKTools, paragon } from '@useparagon/actions'; // Get Paragon Actions to include const actions = await paragon.getActions({ categories: [IntegrationCategory.CRM, IntegrationCategory.PROJECT_MANAGEMENT], }); // Tool calling from available tools await generateText({ model: openai('gpt-4o'), tools: createAISDKTools(actions), toolChoice: 'auto', prompt: 'Help me create a new task in Jira.' });
Expose actions at query-time
Expose actions at query-time
When your AI agent receives a query, it will call the ActionKit API to retrieve available 3rd-party actions and their corresponding parameters.
When your AI agent receives a query, it will call the ActionKit API to retrieve available 3rd-party actions and their corresponding parameters.
Chain and execute actions
Chain and execute actions
Your AI agent will determine (and chain) the appropriate actions based on our agent-specific Actions metadata
Your AI agent will determine (and chain) the appropriate actions based on our agent-specific Actions metadata
Built on an enterprise-ready infrastructure
Built on an enterprise-ready infrastructure
Workflows run in an environment that is built to handle high event volumes, keep user credentials secure and up-to-date, and provide observability and replay features for every request.
Workflows run in an environment that is built to handle high event volumes, keep user credentials secure and up-to-date, and provide observability and replay features for every request.
10M+
Requests Executed Per Day
10M+
Requests Executed Per Day
10M+
Requests Executed Per Day
10M+
Requests Executed Per Day
On-premise
deployment options
On-premise
deployment options
Serverless infrastructure
Serverless infrastructure
SOC 2
Type II
SOC 2
Type II
SSO & Role based access controls
SSO & Role based access controls
Join the ActionKit beta waitlist
We're onboarding new engineering teams every single week for feedback and early testing.
Join the ActionKit beta waitlist
We're onboarding new engineering teams every single week for feedback and early testing.
Join the ActionKit beta waitlist
We're onboarding new engineering teams every single week for feedback and early testing.
Join the ActionKit beta waitlist
We're onboarding new engineering teams every single week for feedback and early testing.
The embedded integration platform for developers
Integrations
© 2024 Paragon Inc. All Rights Reserved
The embedded integration platform for developers
Integrations
© 2024 Paragon Inc. All Rights Reserved
The embedded integration platform for developers
Integrations
© 2024 Paragon Inc. All Rights Reserved
The embedded integration platform for developers
Integrations
© 2024 Paragon Inc. All Rights Reserved