Skip to content
Julius Dela Peña

Selected WorkProject 15AI Agent

FacebookAI Agentwith n8n.

An n8n AI agent that replies to incoming Facebook events, working from a Google Docs reference and conversation memory.

An n8n webhook answers Facebook’s verification check and filters incoming events. An AI agent on Google Gemini, with conversation memory and a Google Docs reference document, prepares each reply and returns it by HTTP request.

  • n8n
  • Google Gemini
  • Google Docs
  • Facebook

15Project
of 15

Project overview
n8n workflow: webhook with verification branch, filter, Google Docs document, AI agent with Gemini model and memory, HTTP requestOpen the full Facebook AI Agent with n8n (opens in a new tab)
01n8n workflow · AI Agent
Open the full Facebook AI Agent with n8n capture (opens in a new tab)
02Close-up
Context
AI Agent
Domain
n8n workflow
Scope
Incoming Facebook event to AI reply
Status
Workflow built in n8n

01The Brief

A reply for every incoming event.

Facebook sends events to a webhook that first has to pass Facebook’s verification check; each incoming event then needs a reply prepared and sent back.

The goal

Verify the webhook, screen incoming events and reply with an AI agent working from a reference document.

n8n workflow: webhook with verification branch, filter, Google Docs document, AI agent with Gemini model and memory, HTTP requestOpen the full n8n workflow · AI Agent capture (opens in a new tab)

Facebook AI Agent with n8n: the full workflow capture. Select it to open full size.

02The Workflow

Verified, filtered, answered.

An n8n webhook answers Facebook’s verification check and filters incoming events. An AI agent on Google Gemini, with conversation memory and a Google Docs reference document, prepares each reply and returns it by HTTP request.

Workflow spec

Trigger
GET requests pass Facebook’s verification check; POST requests carry incoming events.
AI
Gemini agent with memory
Output
HTTP reply

03How It Works

Verify, add context, answer, reply.

Each stage, read straight off the workflow’s own node names.

  1. Webhook

    GET: verify and respond to webhook · POST: incoming events

  2. Context

    Filter · Get a document (Google Docs)

  3. AI agent

    AI Agent · Google Gemini chat model · Simple memory

  4. Reply

    HTTP request

04Architecture

The workflow, start to finish.

The same steps, arranged as they run: in sequence, or split into parallel paths.

  1. 01Webhook

    • TriggerWebhook
    • VerificationGET: verify and respond
      EventsPOST: incoming events

  2. 02Context

    • RulesFilter incoming events
    • Google DocsGet reference document

  3. 03AI agent

    • AI · Google GeminiAI agent
    • Conversation memorySimple memory

  4. 04Reply

    • OutputHTTP request reply (Third-party handoff)
  • Workflow step
  • Third-party platform

The webhook answers Facebook’s verification on GET; POST events are filtered, given the reference document, answered by the Gemini agent and replied to by HTTP.

05Validation & Controls

Where AI runs — and where rules do.

The webhook is verified and events are screened before the agent answers from its reference.

Implementation statusBuilt in n8n

  • WebhookBuilt
  • ContextBuilt
  • AI agentBuilt
  • ReplyBuilt

What each part does

Trigger
Webhook: GET requests pass Facebook’s verification check; POST requests carry incoming events.
Rules
Filter: A filter screens incoming events before the agent sees them.
Context
Google Docs reference: A Google Docs document is loaded as the agent’s reference.
AI
Gemini agent with memory: An AI agent on a Google Gemini chat model, with simple conversation memory, prepares the reply.
Output
HTTP reply: The reply is sent back by HTTP request.

06The Stack

Each tool has a defined job in the workflow.

Roles over logos. Only the tools in the workflow are listed.

  • n8n

    Runs the webhook, filter and agent

  • Google Gemini

    The agent’s chat model

  • Google Docs

    The agent’s reference document

  • Facebook

    Sends events, receives the reply

07The Outcome

What the workflow does.

An n8n AI agent that replies to incoming Facebook events, working from a Google Docs reference and conversation memory.

An n8n webhook answers Facebook's GET verification check and filters incoming POST events; an AI agent on Google Gemini, with conversation memory and a Google Docs reference document, prepares the reply and returns it by HTTP request.

My role

  • Workflow Design
  • Systems Integration

Delivered capability

  • GET requests pass Facebook’s verification check; POST requests carry incoming events.
  • A filter screens incoming events before the agent sees them.
  • A Google Docs document is loaded as the agent’s reference.
  • An AI agent on a Google Gemini chat model, with simple conversation memory, prepares the reply.
  • The reply is sent back by HTTP request.

Contact

Have a workflow, dashboard or CRM that needs fixing? Let’s talk.

Available for new projects

Tell me about the operation, the systems involved and where the work gets stuck. A short call is usually enough to see whether I can help.

Book a call (opens in a new tab)Email me