Lead Gen
Flowise + Google Sheets - monitored by LangSmith
In this integration we're simply going to write the user interaction to a Google Sheet - LeadGen.
Create a folder named: Make
Create a New Google Sheet named: Flowise-Webhook
Rename Sheet 1 to: Notes
Add the column headers: First Name | Surname | eMail | PhoneNumber | Conversation.
Add a test entry:
First Name: Setup
Surname: Ignore
eMail: [email protected]
Phone Number: 1234567
Conversation: test message
Action: Ignore
Make.com
Make.com (formerly Integromat) serves as a powerful workflow automation platform that connects apps and services without requiring complex coding. It excels in creating sophisticated multi-step automation "scenarios" using its visual workflow builder.
A practical use case involves streamlining a lead management process. When a potential customer completes a form on your website, Make.com automatically captures this information, enriches it with additional data from sources like LinkedIn or ClearBit, segments leads based on predefined criteria, creates new contact records in your CRM, assigns appropriate team members, and triggers personalized email sequences—all without manual intervention.
This automation eliminates repetitive tasks, ensures consistent lead processing, reduces human error, and allows your sales team to focus on meaningful customer interactions rather than administrative work. The result is faster response times, improved lead conversion rates, and a more efficient sales process.
Sign up for a Make.com account.
Log In > Templates > Add webhook data to a Google Sheet.
Click on: 'Start guided setup' to create our scenario.
Create a webhook, and give it a name: Flowise-Webhook > Save
Copy: address to clipboard
Why Data Structure Matters with Webhooks
Webhooks transmit data in specific formats (typically JSON), and the receiving system expects this data to be structured in a particular way to process it correctly. Without proper data structure, the webhook payload may be rejected or misinterpreted.
When sending data via webhooks, you need to match the exact field names, data types, and overall structure that the receiving API expects. For example, if a ticketing system requires a "priority" field as an integer (1-5), but you send it as a string ("high"), the webhook request might fail.
Additionally, well-structured webhook data improves debugging and monitoring. When issues arise, having consistent, predictable data formats makes it easier to identify where problems occur in your integration.
Many webhook destinations also implement validation rules that reject improperly structured payloads as a security measure, so adhering to the expected data structure is essential for successful integration.
Postman
Postman is a comprehensive API development and testing platform that helps developers build, test, document, and share APIs. It provides a user-friendly interface that simplifies the often complex process of working with APIs throughout their lifecycle.
At its core, Postman offers a powerful request builder that allows developers to craft HTTP requests with various methods (GET, POST, PUT, DELETE, etc.), add headers, parameters, and request bodies. The platform enables the creation of collections to organize related requests, making it easy to maintain and execute complex API workflows.
Postman's testing capabilities allow developers to write automated tests for API endpoints using JavaScript, ensuring reliability and consistency. The platform also includes features for API documentation, monitoring, mock servers, and team collaboration, making it an essential tool for development teams working with APIs in modern software development environments.
Sign up for a Postman account.
Download and install Postman.
xx
Click on Collections > HTTP
Paste the Flowise-Webhook link and create a POST request
Enter the following:
Body
RAW
Text
JSON
Copy the following:
{
"First Name":"This note was sent by Postman",
"Surname":"Setup",
"eMail":"[email protected]",
"Phone Number ":"1234567",
"Conversation":"test message",
"Action":"Ignore"
}POST the Request. Check for a 200 OK Accepted.
If you go back to your Webhook - Successfully determined
Click: 'Continue'
Ensure you have enabled the Google Sheets API for your account.
Click: Create Connection > Enter Connection Name > Sign into your Google account.
From the drop boxes navigate to the Google Sheet and Sheet Name.
Next .. map the Webhook fields to the Google Sheet - drag & drop fields
Click: 'Continue' to finish.
Click: 'Run the Scenario'.
Go back to Postman and POST the Request.
Check your Google Sheet.
If everything is fine, in Scenarios > Enable Scenario
Finally .. Verify the Connection.
LangSmith
LangSmith is a developer platform created by LangChain to help build, test, evaluate, and monitor language model applications. It provides tools for debugging, optimizing, and managing LLM-powered applications throughout their lifecycle.
LangSmith offers features such as tracing to visualize application execution, dataset creation for systematic testing, human and automated evaluation frameworks, and monitoring capabilities to track performance in production.
Sign up for a LangSmith account.
Sign Up for a Developer account.
Log in > Settings > API Key
Create an API Key
Store the API Key in a safe place ..
In Flowise > Credentials > Langsmith API
Enter the following details:
Open the Chatflow: Simple Chatbot - Lead Gen
Under Settings > Configuration > Analyze Chatflow
Click Save.
In the Chatbot run through the process of booking a call.
Refresh LangSmith
Click on the Project: GS - Lead Gen
LangSmith Lead Generation Project Analysis
This LangSmith dashboard displays monitoring data for a lead generation AI agent project. Over the past week, the system has handled 4 customer conversations with perfect reliability (0% error rate) and full streaming capability.
The agent successfully processed various lead types, including meeting scheduling requests, product inquiries about Pentaho Data Integration, and customer information collection. Performance metrics show reasonable response times, with median latency of 3.62 seconds and first token generation of under 1 second for typical interactions.
Token usage metrics indicate the conversations are efficiently managed, with a total of 6,984 tokens consumed across all interactions and a median of 1,815 tokens per conversation. The system maintains complete conversation traces that can be reviewed for quality assurance.
Last updated