Browse 12,955 Workflows

By Oneclick AI Squad

Quick overview This workflow runs on demand (manual or webhook) to backfill historical data from an HTTP API into Postgres in resumable batches, persisting a checkpoint and progress stats so repeated runs can continue from the last saved position. How it works Starts via Manual Trigger or Webhook Trigger to begin or resume a backfill run. Creates a Postgres backfill_progress table (if needed) and loads the backfill configuration and current checkpoint used to page through history. Checks whether the job is already marked as completed and, if so, skips processing and returns a completion summary. Requests the next batch of records from the source HTTP API using the current checkpoint, end date, and batch limit. Normalizes the API response into mapped fields, determines the next checkpoint value, and proceeds only if the batch contains records. Upserts the batch into the target Postgres table, then saves the updated checkpoint and cumulative progress counters to backfill_progress. Loops with a short wait between batches until the per-run batch limit is reached or no more data is available, then finalizes and outputs a summary (optionally responding via webhook). Setup Create a Postgres credential in n8n and ensure the target table exists (for example, orders_history) with the unique key used for upserts. Update the configuration in the “Code - Load Config & Checkpoint” step with your API base URL, authorization header/token, date range or cursor strategy, field mappings, and target table/unique key. If you want to start the workflow externally, enable the Webhook Trigger and copy the webhook URL into the source system that should initiate the backfill. An n8n automation workflow template by Oneclick AI Squad.

By Oneclick AI Squad

Quick overview This workflow manually ingests data from any paginated HTTP API, normalizes each record, and upserts it into a PostgreSQL table, looping through pages with built-in retries and an optional wait between requests. How it works Starts when you run the workflow manually. Builds an API request configuration (URL, headers, pagination strategy, limits, retry settings) and initializes pagination state. Calls the API via HTTP Request and returns the full response while retrying on failures. Parses the response body, extracts the records from the configured JSON path, maps fields into the target database schema, and computes the next page/cursor/link to request. Upserts the normalized batch into PostgreSQL using an INSERT ... ON CONFLICT statement when the page contains records. Checks whether more pages remain (and the max-pages safety limit is not exceeded), waits briefly to respect rate limits, and repeats the fetch until pagination ends. Outputs a final summary with pages processed and total records ingested, then ends the workflow. Setup Update the API configuration in the code step (base URL, auth header/token, pagination type and parameters, response dataPath/cursorPath, and any static query parameters). Add a PostgreSQL credential, and ensure the target table and unique key column exist and match the configured tableName/uniqueKey and field mappings. Adjust operational limits such as limit, maxPages, maxRetries, retryWaitMs, and the wait delay between pages to fit the API’s rate limits and payload size. An n8n automation workflow template by Oneclick AI Squad.

By Anna Cole

Quick overview This workflow reacts to Stripe dispute events, pulls related WooCommerce order details and recent Gmail messages, uses OpenAI to draft dispute evidence, saves the evidence on the Stripe dispute (as a draft or auto-submitted), and posts updates to Slack while tracking disputes in Google Sheets. How it works Triggers when Stripe sends a charge.dispute.created or charge.dispute.closed event. For new disputes, fetches the dispute and expanded charge from the Stripe API and extracts key payment signals (3D Secure, CVC/AVS checks, customer details, order metadata, and response deadline). Skips disputes that don’t need a response (already submitted, past due, or below the configured minimum amount) and, when an order ID exists, pulls the matching WooCommerce order, the customer’s order history, shipment tracking metadata, and recent customer emails from Gmail. Sends a consolidated, fact-only case file to OpenAI and generates a reason-specific rebuttal draft plus a list of missing evidence. Maps the draft into Stripe’s dispute evidence fields and saves it back to Stripe as a draft, only auto-submitting when autoSubmit is enabled and the case is rated high likelihood. Posts a summary to Slack and appends a row to a Google Sheets “Disputes” tracker, then schedules a Slack reminder before the due date if the evidence remains unsubmitted. For dispute-closed events, updates the dispute outcome in Google Sheets and announces the result in Slack. Setup Create and connect credentials for Stripe (API), WooCommerce (REST API), Gmail, OpenAI, Slack, and Google Sheets. Update the workflow’s configuration values (store name/URL, policy URLs, Slack channel, tracker Google Sheet URL, minimum dispute amount, email lookback window, and reminder timing). Ensure Stripe charges include the WooCommerce order ID in charge metadata (for example order_id from the WooCommerce Stripe Gateway) so the workflow can match disputes to orders. Create a Google Sheets document with a tab named Disputes and columns that match the fields being logged (for example Dispute ID, Order, Customer email, Amount, Reason, Respond by, Recommendation, Win likelihood, Action, Missing evidence, Outcome, Closed on, Stripe link). Requirements Stripe account (works in test and live mode) WooCommerce store using the WooCommerce Stripe Gateway, with REST API keys Gmail account OpenAI API key Slack workspace (optional) Google Sheets with a tab named Disputes (optional) Customization Set autoSubmit to true to send strong cases to the bank automatically Change reminderHoursBeforeDeadline or minimumDisputeAmount in Set Store Configuration Replace Gmail with Zendesk, Gorgias or Help Scout to pull support tickets Swap the OpenAI model or edit the dispute-reason playbook in the AI node Delete the Slack or Google Sheets steps if you don't need them Additional info Evidence is saved to Stripe as a draft by default, so nothing reaches the bank until you review it. Test it in Stripe test mode with card 4000000000000259, which always creates a dispute. The deadline reminder uses a Wait node, so run it on an n8n instance that keeps waiting executions (n8n Cloud or self-hosted with a database). An n8n automation workflow template by Anna Cole.

By Bhaskar Gorati

Quick overview This workflow collects client intake via an n8n Form, uses Anthropic to analyze requirements and draft emails, stores onboarding data in Supabase, creates a Google Drive folder, and sends Gmail and Slack notifications, then follows up until a Supabase checklist is complete. How it works Receives a new submission from an n8n Form with the client’s company, contact details, requested service, and requirements. Normalizes and cleans the submitted fields (for example, trimming text, formatting the website URL, and standardizing the email address). Uses Anthropic to summarize the request, infer priority, and list any missing information needed to start. If required fields are missing, sends the client a Gmail message requesting the missing details and stops the onboarding flow. If the submission is complete, creates a client record and an onboarding checklist in Supabase, creates a client folder in Google Drive, and uses Anthropic to generate a personalized welcome email. Sends the welcome email via Gmail, posts the new onboarding summary to a Slack channel, and updates the client status in Supabase. Waits for the configured delay, checks the Supabase onboarding checklist, and either marks the client as completed and notifies Slack or emails a Gmail reminder listing the remaining checklist items. Setup Connect Supabase credentials and create/configure the clients and onboarding_checklists tables with the fields used by the workflow. Connect Anthropic credentials (via the Anthropic chat model) for requirement analysis and welcome-email drafting. Connect Gmail OAuth2 credentials for sending the welcome and reminder emails. Connect Slack OAuth2 credentials and set the target channel (default is client-onboarding). Connect Google Drive OAuth2 credentials and update the parent “Clients” folder ID used to create each client folder. Set your agency name in the data-cleaning step and adjust the wait duration/checklist items to match your onboarding process. An n8n automation workflow template by Bhaskar Gorati.

By Famulor

Quick overview This workflow runs hourly (or on manual run) to fetch completed negative-sentiment calls from the Famulor API for the last 48 hours, then upserts a concise review row per call into a Google Sheets follow-up queue. How it works Runs every hour on a schedule (or manually for testing). Calculates the creation window in the configured Business timezone, then converts its boundaries to UTC for the Famulor API. Requests completed, negative-sentiment calls from the Famulor Calls API with pagination. Validates the paginated results, deduplicates by Call ID, and formats each call into a short review row (no transcripts/recordings, summary capped at 1,500 characters, and missing success analysis labeled). If no calls match, returns a clear no-results preview message. Otherwise, continues with the prepared review rows. If preview mode is enabled, returns the prepared rows without writing anything; otherwise, upserts the rows into Google Sheets by matching on Call ID. Setup In Famulor, enable sentiment analysis for the assistants you want to monitor and create an API key with calls:read access. In n8n, create an HTTP Header Auth credential with header name Authorization and value Bearer followed by the workspace API key. Select that credential in Read Famulor pages and verify the connection before testing; imported nodes can preselect an existing credential. Connect Google Sheets credentials, set your spreadsheet ID and sheet tab name in the workflow configuration, and create the required headers (for example: Call ID, Created at, Ended at, Direction, Caller, Recipient, Duration minutes, Sentiment, Summary, Success, Review reason). Set Business timezone and Lookback hours in Configure workflow. Keep Preview only enabled for the first run and inspect the prepared rows. After connecting and checking your Google Sheets destination, switch Preview only off and activate the schedule. Requirements A Famulor workspace API key with calls:read, assistants with sentiment analysis enabled, and a Google Sheets account with edit access to the destination spreadsheet. Uses built-in n8n nodes. Customization Adjust Lookback hours, Business timezone and the polling interval. Add Owner, Follow-up status or Notes columns for the team; the workflow leaves these operator columns untouched. Additional info The window filters calls by creation time. Calls or sentiment analysis appearing after the lookback window are not included. Pagination stops at 2,000 records and incomplete scans fail before writing. Google Sheets uses Call ID for upserts; simultaneous executions can still race, so use one active workflow. The destination sheet receives phone numbers and short summaries and should have appropriate access restrictions. API errors stop the execution. An n8n automation workflow template by Famulor.

By Simeon Penev

Quick overview Fact-check AI-written articles before they go live. Type a topic into an n8n form: OpenAI researches and writes the draft as a Google Doc, Lenz checks the factual claims against independent sources, and the form shows the report with Approve, Keep as draft and Edit. How it works A form asks for the topic. An AI agent searches the web with OpenAI's built-in search and writes only from what it found, with sources. If it finds too few good sources, the run stops and the form says so. Lenz Extract (free) lists the factual claims, and Lenz Assess rates them. Red flags get a Lenz Verify deep check. If it confirms an error, the AI rewrites only those sentences. The article becomes a Google Doc with headings, lists and links, and the form shows the report with Approve, Keep as draft and Edit. Before Approve or Edit, the workflow reads the document again and Lenz checks anything you changed by hand. Edit rewrites the article from your note, and Lenz re-checks the changed parts. Setup Install the Lenz node when n8n offers it and add your Lenz API key. Connect Google Docs and Google Drive with your Google account. On self-hosted n8n, add an OpenAI API key to the three AI model nodes. On n8n Cloud they run on n8n credits, so no key is needed. Publish the workflow and open the form's Production URL. Type a topic and keep the tab open. In a few minutes the form shows the report and a link to the document. Requirements A Lenz API key (free plan available) and the Lenz node 0.6.0 or newer, a verified community node for n8n Cloud and self-hosted A Google account with Google Docs and Google Drive On self-hosted n8n: an OpenAI API key (on n8n Cloud the AI runs on n8n credits) Customization In Set your options, change the article length, the minimum number of sources and the Drive folder (My Drive by default). Set how many claims Lenz checks (15 by default) and how many red flags get a deep check (3). Costs: Assess takes about 1 Lenz credit per claim, a deep check 5 credits at Low depth. Additional info Who's it for: content teams, writers and marketers who want the first draft of their AI-assisted blog post they can trust, especially on health and finance topics. An n8n automation workflow template by Simeon Penev.

By Christophe REVOIRE

Quick Overview This workflow runs every Monday at 8am and reads Listings, Leads, and Viewings from Google Sheets to generate a weekly activity summary per seller, then sends each seller an HTML report email via Gmail and updates the listing rows with the last report date. How it works Runs every Monday at 8am on a schedule. Reads listing, lead, and viewing data from the Listings, Leads, and Viewings tabs in a Google Sheets document. Groups active listings by seller email and calculates weekly and total enquiries, viewings, upcoming viewings, average feedback rating, days on market, and recent viewer comments. Builds one HTML email per seller with per-property stats and rule-based advice, and prepares the subject line for the selected reporting period. Sends each seller their report via Gmail, BCC’ing the agent email and using the agency name as the sender name. Writes the current timestamp back to Google Sheets for each reported listing in the “Last Report Sent At” column. Setup Add Google Sheets credentials with access to the target spreadsheet and Gmail credentials for the sending mailbox. Update the Config values, including the Google Sheet ID, agency/agent details, timezone, currency symbol, reportDays, reportStatuses, and staleAfterDays. Ensure your Google Sheets file contains tabs named Listings, Leads, and Viewings with the expected columns (including Seller Email, Listing Ref, Status, Listed At, and Last Report Sent At, plus Viewing feedback fields). Confirm seller email addresses are valid in the Listings sheet and that listing statuses match the values in reportStatuses (for example, available and under offer). An n8n automation workflow template by Christophe REVOIRE.

By Oneclick AI Squad

Quick overview This workflow receives lead submissions via a webhook and uses PostgreSQL to enforce idempotency so the same Idempotency-Key is processed at most once, replaying stored results on retries. For newly claimed requests, it calls external enrichment, CRM, and notification APIs before returning a consistent response. How it works Receives a POST request on a webhook endpoint with an Idempotency-Key (and optional X-Client-Id) plus a JSON lead payload. Ensures the PostgreSQL idempotency_keys table and required indexes exist, then validates the request and generates a canonical fingerprint for the payload. Atomically claims the Idempotency-Key in PostgreSQL with a lease, or loads the existing record if the key already exists. Routes the request to either execute processing (new claim), replay a previously stored completed response, or return an error for in-progress, mismatched payload, or store-unavailable scenarios. For newly claimed requests, sends the lead to an enrichment API, computes a deterministic lead score and tier, then creates/updates the lead in a CRM API and sends a notification via a third HTTP endpoint. Classifies the overall outcome, stores the final HTTP status and response body back to PostgreSQL, and responds to the original webhook with Idempotency-Key and an Idempotent-Replayed flag. Setup Add PostgreSQL credentials and ensure the database user can create tables/indexes and read/write the idempotency_keys table. Replace the placeholder enrichment, CRM, and notification URLs in the workflow configuration (currently pointing to httpbin.org) with your real endpoints. Configure the source system to send POST requests to the webhook URL and include an Idempotency-Key header (and X-Client-Id if you need per-client scoping). Review and adjust policy settings in the workflow configuration such as allowedSources, leaseSeconds, timeouts, maxScoreThreshold, and set allowSimulation=false for production. An n8n automation workflow template by Oneclick AI Squad.

By Oleksii Vasyliev

Quick overview This workflow checks public job postings from Ashby, Greenhouse, and Lever boards plus Jobicy and Himalayas feeds, filters and deduplicates them, scores each new role against your profile with Anthropic Claude, and sends qualifying matches (or unjudged fallbacks) to Telegram. How it works Runs daily at 09:00 (and can be run manually) and loads your settings like target region, profile, keywords, and Telegram chat ID. Fetches job listings from your configured company boards (Ashby, Greenhouse, Lever) and optionally searches Jobicy and Himalayas for the configured keywords. Normalizes all postings into a common format, merges sources, and removes duplicates so each opening is only processed once. Applies low-cost filters for allowed region, matching role titles, and non-junior roles, then keeps only jobs not seen in previous executions and caps the number processed per run. Uses Anthropic Claude (via the LangChain LLM chain) to return a structured verdict (fit, 1–10 score, and a one-line reason) based on your profile, while keeping jobs as “unjudged” if the LLM call fails. Sends a run summary to Telegram and then sends one Telegram message per job that meets the minimum score threshold (or could not be judged), including required “via” attribution links for Jobicy and Himalayas items. Setup Add an Anthropic API credential for the Claude model and a Telegram bot credential for sending messages. Set your Telegram chat ID, profile text (including hard requirements), minimum score, max-per-run limit, and allowed region in the “Configure me” step. Populate the company boards list (company, ATS, slug) and/or enable aggregators with keywords for Jobicy/Himalayas, then run once manually to validate settings before activating the daily schedule. Requirements Anthropic API key Telegram bot token (from @BotFather) and your chat ID Customization Add a company: one line in "Configure me" with its name, ATS (ashby, greenhouse or lever) and board slug Other roles: edit the role filter pattern and the feed keywords Feeds only: empty the company list. Companies only: set useAggregators to false Swap Telegram for Slack or email, or Claude for another chat model. An n8n automation workflow template by Oleksii Vasyliev.

By Oneclick AI Squad

Quick overview This workflow syncs changed customer records from a PostgreSQL “warehouse” table to HubSpot (contact batch upsert) and Segment (identify batch) on a schedule or via webhook, while tracking per-record sync state in Postgres and alerting failures to Slack. How it works Runs every 10 minutes on a schedule or on demand via an HTTP webhook. Loads and validates the sync configuration (source table, key/email columns, destination enablement, and field mappings) and posts a Slack message if the configuration is invalid. Acquires a PostgreSQL run lock to prevent overlapping sync runs. Queries PostgreSQL for rows whose content hash has changed since the last successful sync (or rows due for retry) and exits early by releasing the lock if there are no changes. Builds and sends destination-specific batch requests to HubSpot (CRM contacts batch upsert by email) and/or Segment (identify batch), capturing per-record failures from the API responses. Writes updated per-record sync status (synced/failed with backoff/dead) and run metrics to PostgreSQL, loops through additional batches until drained or the per-run limit is reached, then releases the lock and posts a Slack alert only when failures or dead records occurred. Setup Add PostgreSQL credentials and run the manual setup trigger once to create the state/log/lock tables (and the optional demo table) in your database. Create HubSpot and Segment credentials: HubSpot HTTP Header Auth with Authorization: Bearer and Segment HTTP Basic Auth with your write key as the username and an empty password. Create any custom HubSpot contact properties referenced by your mapping (for example lifetime_value and churn_risk). Update the sync configuration values (warehouse table and columns, enabled destinations, JSON field mappings, batch/retry limits, and Slack incoming webhook URL) and, if using the webhook trigger, copy the n8n webhook URL and call POST /retl-sync-now from your source system when needed. An n8n automation workflow template by Oneclick AI Squad.

By Fafa Modey

Quick Overview This workflow collects a plain-text meeting transcript via an n8n form, sends it to Claude through OpenRouter to generate a structured summary and action items, then prepends the formatted notes to the top of an existing Google Doc stored in Google Drive. How it works Receives a form submission with a single .txt transcript upload plus optional meeting date and meeting title. Validates the file type and size, cleans the transcript text, derives the meeting date and title, and determines whether speaker labels match the configured attendee names. Sends the transcript and meeting context to Claude via the OpenRouter chat completions API and requests a strict JSON response containing agenda, topic bullets, and action items. Validates Claude’s JSON output against word limits, agenda/topic consistency, action-item length, allowed owners, and valid due dates, and stops if any checks fail. Exports the existing Google Doc from Google Drive as HTML, inserts the new meeting notes at the top (keeping older meetings below a divider), and uploads the updated HTML back to the same file. Shows a completion page to the uploader with a link to the updated Google Doc and a count of action items (including unassigned owners). Setup Create a Google Doc to store meeting notes, copy its document ID, and paste it into the notesDocId field in the configuration step. Add a Google Drive OAuth2 credential and select it in the Google Drive export and upload steps. Add an OpenRouter HTTP Header Auth credential with header name Authorization and value Bearer , and select it in the OpenRouter request step. Update the configuration values for organisation name, attendee names, meeting days, timezone, glossary terms, model slug, and transcript/summary limits to match your team. Use the form’s Test URL while building, then activate the workflow and use the Production URL for day-to-day transcript uploads. An n8n automation workflow template by Fafa Modey.

By WeblineIndia

Quick overview This workflow monitors new purchase requisitions in Airtable, uses Groq LLM scoring plus keyword checks and department rules to determine an urgency-based routing decision, logs each decision to Google Sheets, updates the requisition status in Airtable, and notifies the right team in Slack. How it works Triggers when a purchase requisition is submitted in Airtable (based on the submitted_at field). Sends the requisition description to a Groq-hosted LLM to return a strict JSON urgency score (1–5) and short reasoning. Scans the requisition description and business justification for critical/urgent keywords (for example, “outage” or “breach”) and produces a keyword-based urgency score. Looks up department-specific routing rules from a separate Airtable base and calculates a final priority score and routing decision, with cost thresholds forcing VP escalation. Routes the requisition into VP approval, auto fast-track, or the standard queue based on the calculated routing decision. Logs the decision (scores, routing decision, reason, cost, and metadata) to a Google Sheets audit log and updates the original Airtable requisition with the new status and routing note. Sends a Slack message for VP-approval and fast-tracked requisitions to notify the appropriate channel. Setup Connect Airtable credentials (personal access token) and set the correct base/table for both the requisition trigger and the requisition update steps. Create and connect the Airtable base/table that stores department rules (including baseline_criticality, requires_vp_approval_over, and fast_track_eligible) and ensure department names match the requester_department values. Add a Groq API credential for the LLM step and confirm the selected model is available in your Groq account. Connect Google Sheets credentials and set the target spreadsheet and sheet tab used for the append/update audit log. Connect Slack credentials and configure the target channels for VP escalation and fast-track notifications. Additional info How To Customize Nodes Adjust AI Rules:* Open the *LLM: Evaluate Urgency node to modify the scoring guide. For example, you can train the prompt to specifically target software renewals or physical inventory depending on your industry. Modify Critical Keywords:* Open the *Scan: Critical Keywords custom code node. You can easily add or remove specific industry jargon from the criticalKeywords or urgentKeywords arrays. Tweak Math Thresholds:* If you want to change what qualifies as a "Fast Track," open the *Calculate Final Route code node and adjust the finalScore >= 6.0 threshold to a higher or lower number to make the workflow more or less strict. Add‑ons ServiceNow / Jira Integration:** Swap the Airtable trigger for a webhook to ingest purchase requests directly from your IT service management ticketing system. Automated Email Fallbacks:** Add an Outlook or Gmail node parallel to the Slack alerts to ensure executives who are not active on Slack still receive high-cost escalation alerts. ERP Write-Back:** Connect an HTTP Request node at the end of the Fast-Track route to push approved purchase orders directly into SAP, NetSuite, or Workday. Use Case Examples While designed for standard procurement, this triage logic is incredibly versatile. By simply adjusting the Airtable columns and AI prompt, similar use cases include: IT Helpdesk Triage: Automatically scanning IT support tickets to fast-track critical security vulnerabilities while placing password resets in a standard queue. Marketing Budget Approvals: Evaluating ad-hoc marketing spend requests against campaign budgets, fast-tracking small digital asset purchases but escalating major event sponsorships. Maintenance & Facilities: Routing work orders for building repairs, instantly pinging emergency maintenance teams for water leaks while queuing standard lightbulb replacements. Legal Contract Review: Prioritizing incoming vendor contracts based on total contract value and critical keywords (e.g., "liability," "indemnity") to route high-risk agreements to senior counsel. (Note: There can be many more such use cases of this workflow by simply adjusting the data properties and the mathematical constraints!) Troubleshooting Guide | Issue | Possible Cause | Solution | | --- | --- | --- | | Workflow isn't triggering on new requests | Airtable polling issue or wrong field. | Open the Trigger: New PR node. Ensure the Trigger Field is set to the correct date/time column (e.g., submitted_at) and your polling interval is active. | | AI Evaluation fails or returns empty | Groq API rate limit or context window exceeded. | Verify your Groq API credentials. Ensure the Parse: JSON Output node is correctly enforcing the strict JSON schema. | | Routing logic sends everything to the standard queue | Department rules missing or data type mismatch. | Check the Fetch: Dept Rules node output. If it fails to find a matching department name in the secondary table, the fallback math logic will send it to the standard queue. | | Google Sheets audit log is skipping rows | Spreadsheet schema mismatch. | Open the Log: Google Sheets Audit node and ensure the column headers in your actual spreadsheet perfectly match the mapped fields (e.g., PR_ID, Routing_Decision). | Need Help? Building workflows that combine mathematical scoring algorithms, custom JavaScript and advanced AI prompt engineering can be challenging. If you need a helping hand to set up your specific Airtable relational databases, fine-tune the AI intent scoring, or customize Add-Ons like ERP integrations, we are here for you! Our n8n team at WeblineIndia is happy to provide expert assistance with setting up this specific workflow or help you build similar custom enterprise automation processes tailored exactly to your unique business operations. An n8n automation workflow template by WeblineIndia.