Chat with Azure OpenAI Service API on New Contact from Twilio SendGrid API

Pipedream makes it easy to connect APIs for Azure OpenAI Service, Twilio SendGrid and 2,100+ other apps remarkably fast.

Trigger workflow on

New Contact from the Twilio SendGrid API

Next, do this

Chat with the Azure OpenAI Service API

No credit card required

▶

Watch us build a workflow

8 min

Watch now ➜

Trusted by 800,000+ developers from startups to Fortune 500 companies

Developers ♥ Pipedream

Getting Started#

This integration creates a workflow with a Twilio SendGrid trigger and Azure OpenAI Service action. When you configure and deploy the workflow, it will run on Pipedream's servers 24x7 for free.

Select this integration
Configure the New Contact trigger
1. Connect your Twilio SendGrid account
2. Configure timer
Configure the Chat action
1. Connect your Azure OpenAI Service account
2. Configure User Message
3. Optional- Configure System Instructions
4. Optional- Configure Prior Message History
5. Optional- Configure Temperature
6. Optional- Configure N
7. Optional- Configure Stream
8. Optional- Configure Stop
9. Optional- Configure Max Tokens
10. Optional- Configure Presence Penalty
11. Optional- Configure Frequency Penalty
12. Optional- Configure User
Deploy the workflow
Send a test event to validate your setup
Turn on the trigger

Details#

This integration uses pre-built, source-available components from Pipedream's GitHub repo. These components are developed by Pipedream and the community, and verified and maintained by Pipedream.

To contribute an update to an existing component or create a new component, create a PR on GitHub. If you're new to Pipedream component development, you can start with quickstarts for trigger span and action development, and then review the component API reference.

Trigger#

New Contact on Twilio SendGrid

Description:Emit new event when a new contact is created

Version:0.0.6

Key:sendgrid-new-contact

View on GitHub

Twilio SendGrid Overview#

The Twilio SendGrid API opens up a world of possibilities for email automation, enabling you to send emails efficiently and track their performance. With this API, you can programmatically create and send personalized email campaigns, manage contacts, and parse inbound emails for data extraction. When you harness the power of Pipedream, you can connect SendGrid to hundreds of other apps to automate workflows, such as triggering email notifications based on specific actions, syncing email stats with your analytics, or handling incoming emails to create tasks or tickets.

Trigger Code#

import orderBy from "lodash/orderBy.js";
import common from "../common/timer-based.mjs";

export default {
  ...common,
  key: "sendgrid-new-contact",
  name: "New Contact",
  description: "Emit new event when a new contact is created",
  version: "0.0.6",
  type: "source",
  dedupe: "unique",
  hooks: {
    async activate() {
      const currentTimestamp = Date.now();
      const state = {
        processedItems: [],
        lowerTimestamp: currentTimestamp,
        upperTimestamp: currentTimestamp,
      };
      this.db.set("state", state);
    },
  },
  methods: {
    ...common.methods,
    _maxDelayTime() {
      // There is no report from SendGrid as to how much time it takes
      // for a contact to be created and appear in search results, so
      // we're using a rough estimate of 30 minutes here.
      return 30 * 60 * 1000;  // 30 minutes, in milliseconds
    },
    _addDelayOffset(timestamp) {
      return timestamp - this._maxDelayTime();
    },
    _cleanupOldProcessedItems(processedItems, currentTimestamp) {
      return processedItems
        .map((item) => ({
          // We just need to keep track of the record ID and
          // its creation date.
          id: item.id,
          created_at: item.created_at,
        }))
        .filter((item) => {
          const { created_at: createdAt } = item;
          const createdAtTimestamp = Date.parse(createdAt);
          const cutoffTimestamp = this._addDelayOffset(currentTimestamp);
          return createdAtTimestamp > cutoffTimestamp;
        });
    },
    _makeSearchQuery(processedItems, lowerTimestamp, upperTimestamp) {
      const idList = processedItems
        .map((item) => item.id)
        .map((id) => `'${id}'`)
        .join(", ")
      || "''";
      const startTimestamp = this._addDelayOffset(lowerTimestamp);
      const startDate = this.toISOString(startTimestamp);
      const endDate = this.toISOString(upperTimestamp);
      return `
        contact_id NOT IN (${idList}) AND
        created_at BETWEEN
          TIMESTAMP '${startDate}' AND
          TIMESTAMP '${endDate}'
      `;
    },
    generateMeta(data) {
      const {
        item,
        eventTimestamp: ts,
      } = data;
      const {
        id,
        email,
      } = item;
      const slugifiedEmail = this.slugifyEmail(email);
      const summary = `New contact: ${slugifiedEmail}`;
      return {
        id,
        summary,
        ts,
      };
    },
    async processEvent(event) {
      // Transform the timer timestamp to milliseconds
      // to be consistent with how Javascript handles timestamps.
      const eventTimestamp = event.timestamp * 1000;

      // Retrieve the current state of the component.
      const {
        processedItems,
        lowerTimestamp,
        upperTimestamp,
      } = this.db.get("state");

      // Search for contacts within a specific timeframe, excluding
      // items that have already been processed.
      const query = this._makeSearchQuery(processedItems, lowerTimestamp, upperTimestamp);
      const {
        result: items,
        contact_count: contactCount,
      } = await this.sendgrid.searchContacts(query);

      // If no contacts have been retrieved via the API,
      // move the time window forward to possibly capture newer contacts.
      if (contactCount === 0) {
        const newState = {
          processedItems: this._cleanupOldProcessedItems(processedItems, lowerTimestamp),
          lowerTimestamp: upperTimestamp,
          upperTimestamp: eventTimestamp,
        };
        this.db.set("state", newState);
        return;
      }

      // We process the searched records from oldest to newest.
      const itemsToProcess = orderBy(items, "created_at");
      itemsToProcess
        .forEach((item) => {
          const meta = this.generateMeta({
            item,
            eventTimestamp,
          });
          this.$emit(item, meta);
        });

      // Use the timestamp of the last processed record as a lower bound for
      // following searches. This bound will be subjected to an offset so in
      // case older records appear in future search results, but have not
      // appeared until now, can be processed. We only adjust it if it means
      // moving forward, not backwards. Otherwise, we might start retrieving
      // older and older records indefinitely (and we're all about *new*
      // records!)
      const newLowerTimestamp = Math.max(
        lowerTimestamp,
        Date.parse(itemsToProcess[0].created_at),
      );

      // If the total contact count is less than 100, it means that during the
      // next iteration the search results count will most likely be less than
      // 50. In that case, if we extend the upper bound of the search time range
      // we might be able to retrieve more records.
      const newUpperTimestamp = contactCount < 100
        ? eventTimestamp
        : upperTimestamp;

      // The list of processed items can grow indefinitely.
      // Since we don't want to keep track of every processed record
      // ever, we need to clean up this list, removing any records
      // that are no longer relevant.
      const newProcessedItems = this._cleanupOldProcessedItems(
        [
          ...processedItems,
          ...itemsToProcess,
        ],
        newLowerTimestamp,
      );

      // Update the state of the component to reflect the computations
      // made above.
      const newState = {
        processedItems: newProcessedItems,
        lowerTimestamp: newLowerTimestamp,
        upperTimestamp: newUpperTimestamp,
      };
      this.db.set("state", newState);
    },
  },
};

Trigger Configuration#

This component may be configured based on the props defined in the component code. Pipedream automatically prompts for input values in the UI and CLI.

Label	Prop	Type	Description
N/A	`db`	`$.service.db`	This component uses `$.service.db` to maintain state between executions.
Twilio SendGrid	`sendgrid`	`app`	This component uses the Twilio SendGrid app.
	`timer`	`$.interface.timer`

Trigger Authentication#

Twilio SendGrid uses API keys for authentication. When you connect your Twilio SendGrid account, Pipedream securely stores the keys so you can easily authenticate to Twilio SendGrid APIs in both code and no-code steps.

About Twilio SendGrid#

Send marketing and transactional email through the Twilio SendGrid platform with the Email API, proprietary mail transfer agent, and infrastructure for scalable delivery.

Action#

Chat on Azure OpenAI Service

Description:Create completions for chat messages with the GPT-35-Turbo and GPT-4 models. [See the documentation](https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#chat-completions)

Version:0.0.1

Key:azure_openai_service-chat

View on GitHub

Azure OpenAI Service Overview#

The Azure OpenAI Service API provides access to powerful AI models that can understand and generate human-like text. With Pipedream, you can harness this capability to create a variety of serverless workflows, automating tasks like content creation, code generation, and language translation. By integrating the API with other apps on Pipedream, you can streamline processes, analyze sentiment, and even automate customer support.

Action Code#

import azureOpenAI from "../../azure_openai_service.app.mjs";
import common from "../common/common.mjs";

export default {
  ...common,
  key: "azure_openai_service-chat",
  name: "Chat",
  description: "Create completions for chat messages with the GPT-35-Turbo and GPT-4 models. [See the documentation](https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#chat-completions)",
  version: "0.0.1",
  type: "action",
  props: {
    azureOpenAI,
    userMessage: {
      type: "string",
      label: "User Message",
      description: "The user messages to provide instructions to the assistant.",
    },
    systemInstructions: {
      label: "System Instructions",
      type: "string",
      description: "The system message helps set the behavior of the assistant. For example: \"You are a helpful assistant.\"",
      optional: true,
    },
    messages: {
      label: "Prior Message History",
      type: "string[]",
      description: "_Advanced_. Because [the models have no memory of past chat requests](https://platform.openai.com/docs/guides/chat/introduction), all relevant information must be supplied via the conversation. You can provide [an array of messages](https://platform.openai.com/docs/guides/chat/introduction) from prior conversations here. If this param is set, the action ignores the values passed to **System Instructions** and **Assistant Response**, appends the new **User Message** to the end of this array, and sends it to the API.",
      optional: true,
    },
    ...common.props,
  },
  async run({ $ }) {
    const data = this._getChatArgs();
    const response = await this.azureOpenAI.createChatCompletion({
      data,
      $,
    });

    if (response) {
      $.export("$summary", `Successfully sent chat with ID ${response.id}.`);
    }

    const { messages } = data;
    return {
      original_messages: messages,
      original_messages_with_assistant_response: messages.concat(response.choices[0]?.message),
      ...response,
    };
  },
};

Action Configuration#

This component may be configured based on the props defined in the component code. Pipedream automatically prompts for input values in the UI.

Label	Prop	Type	Description
Azure OpenAI Service	`azureOpenAI`	`app`	This component uses the Azure OpenAI Service app.
User Message	`userMessage`	`string`	The user messages to provide instructions to the assistant.
System Instructions	`systemInstructions`	`string`	The system message helps set the behavior of the assistant. For example: "You are a helpful assistant."
Prior Message History	`messages`	`string[]`	Advanced. Because the models have no memory of past chat requests, all relevant information must be supplied via the conversation. You can provide an array of messages from prior conversations here. If this param is set, the action ignores the values passed to System Instructions and Assistant Response, appends the new User Message to the end of this array, and sends it to the API.
Temperature	`temperature`	`string`	What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
N	`n`	`integer`	How many completions to generate
Stream	`stream`	`boolean`	If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message.
Stop	`stop`	`string`	Up to 4 sequences where the API will stop generating further tokens.
Max Tokens	`maxTokens`	`integer`	The maximum number of tokens allowed for the generated answer. By default, the number of tokens the model can return will be (4096 - prompt tokens).
Presence Penalty	`presencePenalty`	`string`	Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Frequency Penalty	`frequencyPenalty`	`string`	Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
User	`user`	`string`	A unique identifier representing your end-user, which can help Azure OpenAI to monitor and detect abuse.

Action Authentication#

Azure OpenAI Service uses API keys for authentication. When you connect your Azure OpenAI Service account, Pipedream securely stores the keys so you can easily authenticate to Azure OpenAI Service APIs in both code and no-code steps.

Before you start, you'll need to deploy a model in the Azure OpenAI Service.

Once that's done, enter the name of your Azure OpenAI resource, the deployment name you chose when you deployed the model, and your Azure OpenAI key below.