AI Automation
923 results — all source-linked n8n references
SRun a voice-enabled AI assistant on Telegram with OpenAI, Gmail and Notion
Quick overview This workflow turns a Telegram bot into a personal assistant that handles text and voice messages using OpenAI, can read Gmail and Google Calendar, looks up tasks and contacts in Notion, and can draft and send emails with a required approval step in Telegram. How it works Triggers when a new Telegram message is received by your bot. Detects whether the message is a voice note or text, and if it is voice, downloads it from Telegram and transcribes it with OpenAI. Sends the resulting user text to an OpenAI-powered agent that keeps a short conversation memory per chat. When needed, the agent fetches recent emails from Gmail, lists upcoming events from Google Calendar, and pulls tasks or contacts from Notion to answer the request. If the agent decides to send an email, it asks for approval in Telegram and only proceeds after you approve within the time limit. Sends the assistant’s final response back to the same Telegram chat. Setup Create a Telegram bot with @BotFather, add the Telegram credentials in n8n, and use the workflow’s Telegram trigger to connect the bot. Add an OpenAI API credential for both chat responses and voice transcription, and set your preferred model
Generate daily auto-service AI videos and post them with OpenAI, Perplexity, AtlasCloud, and Blotato
Quick overview This workflow runs daily (or manually) to research an auto-service topic with Perplexity, generate a caption and text-to-video prompt with OpenAI, render an AI video via AtlasCloud, then publish it to TikTok, YouTube, Instagram, and Facebook through Blotato while sending a Telegram preview. How it works Runs every morning on a schedule (or starts via a manual trigger). Loads your business settings (name, topic, language, video format, and CTA) from a central configuration. Uses Perplexity web research plus an OpenAI chat model to produce a structured caption, hashtags, and a cinematic text-to-video prompt. Sends the prompt to AtlasCloud to generate a short vertical video and waits before polling AtlasCloud for the render status. Repeats the wait-and-check loop until AtlasCloud reports the video render as completed. Prepares the final caption and video URL, then publishes the video to TikTok, YouTube, Instagram Reels, and Facebook Reels via Blotato and sends a Telegram message with the post preview and video link. Setup Add credentials for OpenAI, Perplexity, AtlasCloud (HTTP Header Auth), Blotato, and Telegram. Update the Configuration values (business name, topic, C
Create AI product video ad Shorts from Telegram photos with OpenAI and AtlasCloud
Quick overview This workflow takes a product photo sent via Telegram, uses OpenAI Vision to describe it and OpenAI to write a short-form video ad prompt, generates a vertical video via AtlasCloud, then (after two approvals in Telegram) publishes the Short to YouTube via Blotato. How it works Triggers when a user sends a product photo (optionally with a caption/campaign) to a Telegram bot. Downloads the image from Telegram and uses OpenAI Vision to extract an advertising-focused product description. Uses OpenAI to generate a single cinematic text-to-video prompt plus a YouTube Shorts title and description. Sends the proposed prompt to Telegram and waits for your approval before proceeding. If approved, requests video generation from AtlasCloud and polls the render status until the video is completed. Sends the generated video back to Telegram for preview and waits for a second approval to publish. If approved, publishes the video to YouTube as an unlisted Short via Blotato and notifies you in Telegram (otherwise it stops and sends a decline message). Setup Create and connect a Telegram bot credential, then start a chat with the bot so it can receive photos and send approval prompts.
Publish AI social posts from Telegram to LinkedIn, Instagram, Facebook and Threads with OpenAI and Blotato
Quick overview This workflow listens for a Telegram message or voice note, uses OpenAI to turn it into a cross-platform social caption and matching AI image, publishes the post via Blotato to LinkedIn, Instagram, Facebook, and Threads, then sends a confirmation back to Telegram. How it works Triggers when a new Telegram message is received by your bot. Routes the input as voice or text, downloading voice notes from Telegram and transcribing them with OpenAI. Sends the resulting text to an OpenAI chat model to generate a single ready-to-publish caption plus an image prompt. Generates an image with OpenAI (gpt-image-1) using the returned image prompt. Uploads the generated image to Blotato to get a hosted media URL. Publishes the caption and image to LinkedIn, Instagram, Facebook Pages, and Threads through Blotato. Sends a Telegram message confirming publication and includes the posted caption. Setup Create a Telegram bot in @BotFather, add the bot token as a Telegram credential in n8n, and activate the Telegram Trigger. Add an OpenAI API credential for audio transcription, caption generation, and image generation, and (optionally) adjust the model, language, image size, and quality
WSend weekly blended ad performance reports with Meta, Google Ads, GA4 and OpenAI
Quick overview This workflow runs every Monday at 8am to pull the last 7 days of performance from Meta Ads, Google Ads, and GA4, calculate blended KPIs, generate an OpenAI-powered narrative summary, email an HTML report via Gmail, and post an internal update (and errors) to Slack. How it works Runs every Monday at 8am on a scheduled trigger. Fetches the last 7 days of spend and engagement metrics from the Meta Ads Insights API, campaign cost and conversion metrics from the Google Ads API, and conversions and revenue by channel from the GA4 Reporting API. Combines the three datasets and calculates blended KPIs including total spend, total conversions, total revenue, and ROAS. Sends the computed metrics to an OpenAI chat model to generate a 3–4 sentence client-friendly performance summary with one risk and one recommendation. Builds a simple HTML report containing the KPIs and AI summary. Emails the HTML report to the client via Gmail and posts a brief confirmation with spend and ROAS to a Slack channel. If the workflow fails, posts an error alert message to an ops Slack channel. Setup Configure authentication for the HTTP requests to Meta Ads (access token), Google Ads (OAuth/access
Manage Google and Trustpilot reviews with OpenAI and Slack
Quick overview This workflow runs hourly to fetch new Google Business Profile and Trustpilot reviews, deduplicates them, uses OpenAI via a LangChain agent to draft sentiment-aware replies, escalates negative reviews in Slack, and routes positive replies to Slack for approval before posting the reply back to Trustpilot. How it works Runs every hour on a schedule. Fetches the latest reviews from the Google Business Profile API and the Trustpilot API and combines the results. Deduplicates reviews by tracking previously seen review IDs in workflow static data and stops if there are no new reviews. Processes each new review and uses OpenAI (via a LangChain agent) to generate structured JSON containing sentiment, a draft reply, and an optional suggested resolution. Sends negative reviews to a Slack channel for the owner along with the suggested resolution. Sends positive draft replies to Slack for one-tap approval (with a 24-hour timeout) and, if approved, publishes the reply to the Trustpilot API. Sends a Slack alert to the ops channel if the workflow errors. Setup Configure authentication and replace placeholders for the Google Business Profile reviews endpoint (account and location) i
WHandle WhatsApp support chats with OpenRouter, Pinecone, and Gemini
Quick overview This template implements a WhatsApp support suite that logs inbound events to a dashboard API, routes conversations through an OpenRouter-powered AI agent with Pinecone RAG and memory, exposes a webhook for human outbound replies, and provides a webhook to summarize recent chats for handoff. How it works Triggers on WhatsApp Cloud API events and routes status updates (sent/delivered/read) to a dashboard API endpoint for storage. For inbound messages, looks up the contact in the dashboard API and normalizes the message into a consistent schema (sender, type, timestamp, and best-effort content). If the message contains media (image/video/audio/document), fetches the WhatsApp media URL, downloads the file, uploads it to the dashboard’s media endpoint, and attaches the resulting media URL and MIME type. Stores inbound messages and reactions in the dashboard API, then checks via the cases endpoint whether the sender already has an open case. If the inbound message is text and no open case is found, queries Pinecone as a tool (using Google Gemini embeddings), uses an OpenRouter chat model with conversation memory to draft a reply or create a new case via the dashboard API,
Confirm and log WhatsApp restaurant orders with Groq and Google Sheets
Quick Overview This workflow triggers on incoming WhatsApp orders, uses Groq to extract items and quantities, validates them against a live Google Sheets menu, logs the result to an orders sheet, and replies to the customer with either a clarification request or a priced confirmation (with optional VIP owner alerts). How it works Triggers when a new WhatsApp message is received on your WhatsApp Business Cloud number. Sends the message text to Groq (Llama 3.3) to extract a structured list of ordered items and quantities. Reads the latest menu (item name, price, availability) from a Google Sheets “Menu” tab. Matches extracted items to the menu, checks availability, calculates line subtotals and the total price, and compiles any validation issues. If there are issues, replies to the customer on WhatsApp asking for a corrected order and appends a “needs_clarification” entry to the Google Sheets “Orders” tab. If the order is valid, appends a “confirmed” entry to the Google Sheets “Orders” tab and sends a WhatsApp confirmation with the item summary and total. If the confirmed total meets or exceeds the VIP threshold, sends a separate WhatsApp notification to the owner number. Setup Conne
Predict Salesforce deal escalation risk with Gemini and Slack alerts
Quick overview This workflow pulls open Salesforce opportunities and their last-30-days activities, uses Google Gemini to predict escalation risk per deal, then flags high-risk opportunities in Salesforce and notifies a Slack user. How it works Starts manually and fetches open opportunities and recent Salesforce Task activities from the last 30 days. Merges opportunities with their related activities and builds a per-deal context including amount, probability, type, activity list, and whether all activities are completed. Batches the deals and skips ones that have 100% probability and active/completed engagement, then waits briefly between remaining items to reduce API load. Sends each deal context to Google Gemini to classify risk (high/medium/low), identify specific risk flags, and return a one-line summary as JSON. Cleans and validates the Gemini response by stripping formatting, parsing JSON safely, and normalizing invalid risk levels to “unknown.” For deals marked as high risk, updates the Salesforce opportunity description with the risk details and sends a formatted alert message to a specified Slack user. Setup Add Salesforce OAuth2 credentials and ensure the workflow has ac
Detect revenue leakage with Salesforce, Google Sheets, Gemini, and Slack
Quick overview This workflow manually audits Closed Won Salesforce opportunities against billed amounts in Google Sheets, uses Google Gemini to analyze any revenue gap, then alerts a Slack user and writes an audit note back to the Salesforce opportunity. How it works Starts when you run the workflow manually. Fetches up to 20 Salesforce opportunities where the stage is "Closed Won" and processes them in batches with a short delay to reduce API pressure. Looks up matching billing rows in Google Sheets and totals the billed amount for the same Deal ID, then calculates the missing revenue versus the Salesforce Amount. For deals with a positive revenue gap, sends the discrepancy details to Google Gemini to classify severity and generate a root cause, recommendation, and Salesforce-ready audit note. Parses the Gemini JSON response and builds a formatted Slack message containing the deal, gap, severity, and recommended action. Sends the Slack alert and updates the Salesforce opportunity description with the audit note and the expected, billed, and missing amounts. Setup Add credentials for Salesforce OAuth2, Google Sheets OAuth2, Google Gemini (PaLM) API, and Slack OAuth2. In Salesforce,
Monitor ride surge events and analyze ride errors with Sheets, Slack and Gemini
Quick overview This workflow pulls ride events and error records from Google Sheets, detects surge pricing and long waits, and notifies via Slack. For error records, it uses Google Gemini with real-time traffic and weather API lookups to diagnose the issue and send recommended actions to Slack. How it works Runs manually and reads ride records from a Google Sheets spreadsheet. Processes the rows one by one with a short delay and routes each record based on whether its type is an event or an error. For event records, checks if the ride price is more than 1.5× the normal price and the wait time is over 10 minutes. If surge conditions are met, sends a surge and delay notification to Slack and records a surge-alert log entry. If surge conditions are not met, records a normal “no alert needed” log entry. For error records, structures the ride and error context and asks Google Gemini to generate an incident summary, probable cause, and recommended actions, calling traffic and weather APIs as needed. Parses the Gemini response into JSON and posts the formatted error report to Slack. Setup Connect Google Sheets OAuth credentials and set the correct spreadsheet and sheet tab in the Google S
Generate executive decision briefings from APIs with OpenAI and Google Sheets
Quick overview This workflow runs daily or via an on-demand webhook to pull financial metrics, operational KPIs, and market quotes via HTTP APIs, then uses OpenAI to detect anomalies and generate an executive briefing that is POSTed to a dashboard endpoint and logged to Google Sheets. How it works Runs on a daily schedule or receives a POST request via a webhook to start an executive briefing run with a configurable lookback period. Fetches financial metrics, operational KPIs, and market data in parallel from three HTTP API endpoints and continues even if one source fails. Merges and normalizes the three datasets into a single snapshot, marking missing sources as unavailable so downstream steps keep a consistent structure. Uses OpenAI to analyze the snapshot for anomalies, unusual trends, and data gaps, returning a severity-scored findings JSON. Uses OpenAI again to turn the anomaly findings and raw metrics into a structured executive briefing with a health score, risks, opportunities, and recommended actions. Formats a final JSON payload and delivers it by POSTing to your dashboard endpoint and appending a summary row to a Google Sheets tracker. If the run was triggered by the web
Guard Hostinger VPS CPU spikes with Anthropic, SSH, and email approvals
Quick overview This workflow monitors a Hostinger VPS every 15 minutes for CPU spikes, correlates them with traffic anomalies, and either auto-activates a lockdown firewall or emails a process report with AI analysis and an email-reply approval path to snapshot and kill high-CPU processes. How it works Runs every 15 minutes and loads the VPS ID, SSH host, alert email, and CPU/traffic thresholds. Fetches the last two hours of CPU and network traffic metrics from the Hostinger API and determines whether the latest CPU usage exceeds the configured threshold and whether traffic is anomalous. When a CPU spike is detected, connects over SSH to run ps aux --sort=-%cpu, formats the top processes, and flags PIDs exceeding the kill CPU threshold. Sends the raw process list to Anthropic (Claude) to generate a short diagnosis and recommendation. If the spike appears traffic-driven, lists/creates a Hostinger “lockdown” firewall that only allows the configured ports, activates it on the VPS, and emails a lockdown alert. If the spike is not traffic-driven, writes a pending-approval case entry to /var/log/n8n-cpu-guard.log over SSH and emails a report with the case ID, process table, AI analysis,
Route meeting action items to Slack using Gmail, OpenAI, and Google Sheets
Quick overview This workflow watches a Gmail label for new meeting transcript emails, extracts the transcript, uses OpenAI to generate a summary and action items, routes urgent items to the right Slack channel, and logs the meeting and action items to Google Sheets. How it works Triggers every minute when an unread email arrives in Gmail with the label Transcripts. Applies your configured Google Sheets and Slack settings, then extracts meeting metadata and transcript text from the email body. Reads the existing meeting log from Google Sheets and skips processing if the email matches a previously logged message ID or meeting hash. Sends the transcript to OpenAI to produce a JSON analysis including summary, decisions, risks, and structured action items. Parses and normalizes the action items, then routes the result as urgent, standard, or FYI based on priority and due date threshold. Posts a formatted message to the appropriate Slack channel when action items exist, otherwise marks the run as FYI. Appends the meeting record to the Meetings sheet and writes each extracted action item as a separate row in the ActionItems sheet. Setup Connect credentials for Gmail OAuth2, Google Sheets
Create TikTok carousel posts with Upload Post, Google Sheets, Gemini, GPT-4.1
Quick overview This workflow runs daily to pull TikTok post and profile analytics from Upload Post, updates a Google Sheets performance log, uses Google Gemini and OpenAI to generate a new carousel concept and caption, creates four images via fal.ai, publishes to TikTok, and updates a persistent “Agent Skill” sheet. How it works Runs on a daily schedule (or manually) and loads product, Upload Post, and Google Sheets configuration values. Reads the Google Sheets “Performance Log,” selects posts missing stats or posted recently, fetches TikTok post analytics from the Upload Post API, and writes updated metrics back to the sheet. Aggregates the refreshed history, fetches TikTok profile analytics from Upload Post, reads the “Agent Skill” sheet, and compiles a single context summary. Uses Google Gemini (via a structured JSON output) to generate a TikTok carousel plan including caption, strategy notes, four slide prompts, hashtags, and an engagement question. Sends the four prompts to fal.ai to generate four 9:16 PNG carousel images and collects their hosted image URLs. Uses OpenAI (GPT-4.1) to write a short TikTok photo-post description, publishes the carousel to TikTok via Upload Post
Extract invoice data from Google Drive images with GPT-4o and Postgres
Quick overview This workflow watches a Google Drive folder for new invoice images, extracts key invoice fields with OpenAI vision, checks for duplicates in Postgres, and then stores the invoice or moves the file into processed, duplicate, or failed folders. How it works Triggers every hour when a new file is created in a specific Google Drive folder. Downloads the new file from Google Drive and sends the image to OpenAI (GPT-4o) to extract invoice fields as JSON. Cleans and parses the AI response into structured invoice data, routing invalid or unparseable results to a failed path. Checks Postgres to see if an invoice with the same vendor name, invoice number, and total already exists. If the invoice is a duplicate, moves the source file in Google Drive to a duplicate folder. If the invoice is not a duplicate, inserts the invoice into the Postgres invoices table (status set to processed) and moves the source file to a processed folder. Setup Connect Google Drive OAuth2 credentials and set the incoming, processed, duplicate, and failed folder IDs to match your Drive structure. Add an OpenAI API credential and ensure the selected model (gpt-4o) is available for image analysis. Add Po
C
Post TikTok, YouTube Shorts and Instagram Reels with Airtable, OpenAI and Google
Quick Overview This workflow runs on a schedule, pulls the next publish-ready video from Airtable, uses OpenAI, Google Gemini, and CloudConvert to generate a transcript, title, description, tags, and hashtags, then uploads to YouTube Shorts, posts an Instagram Reel via Meta Graph API, and queues TikTok publishing via Google Sheets. How it works Runs on a schedule (or manually) and waits a random number of minutes to stagger posting. Searches Airtable for content marked Publish = true and Status = “In Publish Queue”, then selects a single record to process. Downloads the video, converts it to MP3 with CloudConvert, and transcribes the audio with OpenAI. Uses OpenAI and Google Gemini to extract a main keyword, generate a YouTube-optimized title, and create YouTube tags, hashtags, and a full YouTube description, then derives a shorter content-only caption for Instagram/TikTok. Uploads the video to tmpfiles.org to generate a direct file URL and updates the Airtable record with the generated metadata and set Status = “Published”. Uploads the video to YouTube with the generated title, description, and tags. Creates and publishes an Instagram Reel via the Meta (Facebook) Graph API after a
WScore LinkedIn job matches with OpenAI, Google Sheets, Slack, and Apify
Quick overview Runs fetch_cat/linkedin-jobs-scraper daily to collect recent LinkedIn jobs, scores each listing against your candidate profile with OpenAI, saves qualified matches to Google Sheets, and sends the five strongest results to Slack without redelivering previously processed job IDs. How it works Runs on a daily schedule (or via manual trigger) and ensures an n8n Data Table exists to track which LinkedIn job IDs have already been delivered. Runs the FetchCat LinkedIn Jobs Scraper (fetch_cat/linkedin-jobs-scraper) through Apify's API using your configured keywords and location, waits for the Actor run to finish, and downloads up to 10 structured job records from its dataset. Normalizes the scraped job records, caps the list to 10 items, and skips any jobs that already exist in the delivery ledger. Sends the remaining jobs to OpenAI in one structured request to score each job against your candidate profile and filters out results below the minimum score. Appends or updates qualified matches in Google Sheets (upserting by LinkedIn job ID) and formats a Slack message containing the five highest-scoring jobs. Posts the digest to your chosen Slack channel and records the deliver
SRun a Telegram voice and chat assistant with OpenAI, Gmail, Calendar, and Notion
Quick overview This workflow turns a Telegram bot into a personal assistant that understands text or voice notes (transcribed with OpenAI), can read Gmail and Google Calendar, look up Notion tasks and contacts, and can draft and send emails with an approval step in Telegram. How it works Triggers when a new Telegram message arrives. Loads assistant settings (name, OpenAI model, languages, memory window) and checks whether the message is a voice note or plain text. If it is a voice note, downloads the audio from Telegram and transcribes it with OpenAI; otherwise it uses the message text as the user input. Sends the user input to an OpenAI-powered agent that keeps short-term chat memory and decides whether to call Gmail, Google Calendar, or Notion tools to answer the request. If the agent decides to send an email via Gmail, it asks for explicit approval in Telegram before proceeding. Sends the agent’s final response back to the user in Telegram. Setup Create a Telegram bot with @BotFather, add your Telegram credentials in n8n, and start a chat with the bot so it can reply to you. Add an OpenAI API credential and set the desired chat model and transcription language in the Configurati
Create and approve AI social posts with OpenAI, Telegram and Blotato
Quick overview This workflow runs daily to generate an AI-written social post and DALL·E image with OpenAI, sends it to Telegram for approval, and—if approved—publishes it to LinkedIn, X, Instagram, Facebook, and Threads via Blotato while logging outcomes to Google Sheets. How it works Runs every 24 hours on a schedule trigger. Uses OpenAI to generate a single content topic (title, angle, and hook) based on your niche, audience, and language settings. Uses OpenAI to write one social post with a separate hashtag line and an image prompt in your specified tone and style. Uses OpenAI Images (DALL·E) to generate a matching image and extracts a shareable image URL. Sends the post text, hashtags, and image link to Telegram and waits for you to approve or skip within the configured time window. If approved, publishes the post with the image to LinkedIn, X, Instagram, Facebook, and Threads using Blotato and appends a “Published” log row to Google Sheets. If skipped or not approved, appends a “Skipped” log row to Google Sheets. Setup Add OpenAI API credentials and ensure the selected chat model and image model in the configuration match models available to your account. Add Telegram credent
WTrack and analyze TikTok videos with Apify, OpenAI, and Google Sheets
Quick overview This workflow runs daily (or manually) to fetch recent videos from a list of TikTok accounts via Apify, transcribes each video with OpenAI, generates content insights and translations, and then appends or updates one row per video in Google Sheets. How it works Runs on a daily schedule (or manually) and loads configuration values like the date window, video limit, output language, and Apify actor. Reads the TikTok usernames to monitor from the accounts tab in Google Sheets. Calls the Apify TikTok Scraper API to retrieve the latest videos for each configured account. Filters and normalizes the results to keep only videos published within the last N days and extracts metrics, captions, hashtags, and download URLs. Downloads each video file and transcribes it with OpenAI speech-to-text. Sends the transcription plus caption and hashtags to OpenAI to extract a hook, classify the content, infer the video format, and translate the transcription. Writes the combined video metadata, transcription, and AI insights to the insights tab in Google Sheets using an append-or-update operation. Setup Create a Google Sheets spreadsheet with an accounts sheet containing a username colum
Generate CCTV-style animal TikTok videos with OpenAI, Seedance and Blotato
Quick Overview This workflow runs every four hours to generate a CCTV-style animal video concept with OpenAI, renders an 8-second vertical clip using AtlasCloud Seedance, then publishes it to TikTok via Blotato and sends a preview to Telegram. How it works Runs every 4 hours on a schedule. Uses OpenAI to create a hyper-realistic CCTV-style image and video prompt based on a random (or overridden) animal, location, action, and a current timestamp overlay. Starts a Seedance text-to-video render on AtlasCloud with the generated prompt, duration, aspect ratio, and optional audio. Polls AtlasCloud at a fixed interval until the render status is completed or succeeded and retrieves the output video URL. Uses OpenAI to generate a TikTok-ready title and exactly five hashtags from the video concept. Publishes the generated video to TikTok via Blotato with the title/hashtags and the selected privacy level, flagged as AI-generated. Sends the rendered video and title as a preview to a specified Telegram chat. Setup Add OpenAI API credentials and set the prompt and caption models in the Configuration values if you want to change them. Add an AtlasCloud API key (HTTP Header Auth) and ensure it is
Route WhatsApp support requests with Claude, Supabase, HubSpot and Slack
Quick overview This workflow turns incoming WhatsApp messages (text, audio, images, and documents) into actionable support interactions by extracting content, detecting sentiment, and using an AI agent with a Supabase knowledge base to route requests to Google Calendar, HubSpot, Slack, Gmail, and back to WhatsApp. How it works Receives incoming WhatsApp messages via a WhatsApp webhook trigger. Routes the message by type and converts it into text using OpenAI Whisper for audio, OpenAI GPT-4o Vision for images, or a PDF extractor for documents (or passes text messages through). Loads the sender’s session context from Supabase and runs sentiment/language detection on the extracted text. Uses a LangChain AI agent (Claude) with Supabase vector search (with OpenAI embeddings and Cohere reranking) plus tool access (Google Calendar, HubSpot, Slack, and Gmail) to produce a structured intent, response, and metadata. Logs conversation analytics (intent, sentiment, message type, and response length) to an n8n Data Table. Routes by intent to either create a Google Calendar event, create/update a HubSpot contact, email a document via Gmail (and upload it to Google Drive), create a HubSpot ticket
Send prescription PDFs to patients via WhatsApp with Google Gemini and Supabase
Quick overview This workflow listens for incoming WhatsApp messages from clinic staff, uses Google Gemini to extract prescription details, generates a prescription PDF via an HTML-to-PDF API, uploads it to a public URL, sends the document back over WhatsApp, and logs outcomes to Supabase. How it works Triggers whenever a new WhatsApp message is received on your WhatsApp Business Cloud number. Captures configuration values and the incoming message text and sender number. Uses Google Gemini to extract patient name, doctor name, and medication details into structured data and flags any missing required fields. If required details are missing, replies to staff on WhatsApp with what to fix and inserts a needs_review record into Supabase. If details are complete, renders a prescription HTML document, converts it to a PDF via an external HTML-to-PDF API, and uploads the PDF to an external file-hosting API. Sends the prescription PDF to the sender on WhatsApp and inserts a “sent” log record (including the PDF URL) into Supabase. Setup Connect WhatsApp Business Cloud credentials for the WhatsApp trigger and the WhatsApp send steps, and configure the workflow’s webhook in Meta/WhatsApp so in