Akogun ImoleayoVerified StackedHQ member

I built an AI email draft workflow that attaches necessary files and sends them to an independent consultant (my client) for them to review. It is an 8-node workflow.

I built a session confirmation and follow-up workflow for a tutorial agency. The workflow checks for sessions a day before the actual session and sends a WhatsApp message to the parents of their tutees to confirm the availability of their wards and also sends a follow-up message a day after to check up and also know if they would like to book another session. These all get updated on Google Sheets, and the tutorial agency gets a notification of messages sent. Tool stack: Make.com, Google Sheets, whatsapp api, telegram

Reads inbound customer inquiries from Gmail and a web form. Extracts action, deadline, preferred time, and open questions using an LLM (Groq Llama 3.3), then routes the result to team email (the CEO, in this case) and a Google Sheets log. Stack: n8n, Groq, Gmail node, Webhook, Google Sheets. Flow: Gmail Trigger + Form Webhook → Normalize (unify field names across both sources) → Merge → HTTP Request to LLM → Parse JSON (with a fallback for malformed output, flagged urgent for manual review) → team email + sheet log. Confirmation branch: looks up the exact row by a unique ID, checks status, sen

I built an automated lead intake system for a marketing agency using n8n, Tally, OpenAI, and Google Sheets. It solves the problem of losing leads due to the massive compilation of forms submitted. This workflow does the heavy lifting of reading the forms, extracting necessary info, storing the info and alerting the person in charge of leads. It eliminates manually reading each submission, which takes time and can lead to lost leads. The flow runs like this: Tally form submission triggers n8n, OpenAI turns the raw message into a one-line summary, the lead gets logged to Google Sheets with a tim

I built a reminder engine for a travel agency. What the workflow does is, every morning by 6am, it checks through a Google Sheet containing client info such as their WhatsApp number, destination, departure date and return date; using a filter (status=pending), it calculates the date for departure (1-3 days before) and the return date (1 day after); and it drafts up a message, which is then sent to the client based on whether it is pre- or post-trip. Then, to prevent duplicate messages being sent, the engine auto-updates the Google Sheet status from pending to pre- or post-trip, based on whic

Built two mini AI workflows, using Gmail, Google Sheets, GroqAI, Google Docs API and Telegram. Email is received, summarised and logged into a Google Sheet, then a draft is made and documented, and the link lands in the Teams Telegram account.

Built an email tracker that summarises and logs each summary from donors into a Google Sheet. Then a second workflow picks the logged information and creates a draft using a given template and notifies the team to review the document. Stack: Make, Gmail, Google Sheets, Google Docs, GroqAI, Google Docs and Telegram.

I build a lead caption, prioritisation and outreach engine that receives prospective leads, classifies the lead into hot, warm and cold, sends dedicated emails to each, alerts the team and enters the information and request into a CRM tool. It also has an error handler that ensures that the workflow does not stop and the team is alerted of every error encountered. Stack: Webhook, Fillout form, HubSpot, Slack, Gmail.

Built a workflow that handles the boring, repetitive and mind-numbing job of handling support tickets manually. The workflow handles support emails received by an e-commerce shop, drafts a response based on the company's policy and drops it in the email of a team member. If the response to a support ticket isn't in the policy doc provided, it escalates and sends a Slack alert to the team. All these get logged to the CRM (Airtable). Approximately 30 mins/request saved. Stack: Gmail, Google Docs, Slack, Airtable, OpenAI, Make.com

The problem Small SaaS teams answer the same support questions on repeat. Someone reads each incoming email, copies the answer out of a help doc, sends the reply, then logs the ticket in the CRM by hand. It's repetitive, slow, and doesn't scale as ticket volume grows. What I built An n8n workflow that reads incoming support emails, checks them against an FAQ knowledge base, and drafts a reply using AI; an actual person still reviews and sends it. Nothing goes out automatically. If the AI isn't confident the FAQ covers the question, it skips drafting and flags the email for a human instead of g

Quick Lead Qualify & Team Alert An n8n workflow that automatically captures form submissions, uses AI to score each lead as hot, warm, or cold and tag the project type, logs everything into Airtable, then instantly notifies your sales team via Slack for hot leads and sends the prospect a personalized email confirmation all within seconds of form submission, no manual triage required.

Missed-Call Text-Back Automation, Bright Smile Dental Built a missed-call detection and follow-up system using MacroDroid (Android automation), n8n, and Google Sheets. Twilio substitution: Twilio trial accounts don't support Nigerian number provisioning, blocking the intended stack. Rebuilt the call-detection layer using MacroDroid's native Missed Call trigger instead, which fires locally on-device and requires no telecom API or paid account. Flow: MacroDroid detects a missed call, sends caller number, timestamp, and status to an n8n webhook (POST, JSON) n8n normalizes the timestamp and lo

I built an Ai Image Cleaner Stack: n8n, Dropbox API, remove.bg API, Google Sheets API, Slack, Telegram The workflow watches a Dropbox /raw-uploads folder and turns raw product photos into store-ready images without any manual editing. The workflow polls the folder every 3 minutes with a Schedule Trigger, lists the contents, and diffs it against a list of already-processed file IDs stored in workflow static data. That stops the same photo from getting reprocessed on the next poll. Each new file goes through a filter that checks the extension before anything else runs. Valid images get logged in

A cross-department notification and approval router in n8n that ingests events from three different tools, Airtable, Slack, and Google Sheets, normalizes them into one consistent shape, then uses AI to classify each item by department (ops, sales, or support) rather than relying on hardcoded rules per source. A Switch node routes each event dynamically based on that classification. Items needing sign-off get flagged and routed separately into an approval path. Every event, regardless of source or department, gets logged as a task in one shared tracking sheet, so managers have a single view of

What it does: an n8n workflow for a stationery supplier that turns incoming email quote requests into logged, priced, staff-reviewed draft replies with no manual data entry. Pipeline: Gmail trigger catches new mail → a classification pass filters out anything that isn't a genuine new quote request → a duplicate check against thread ID stops repeat logging on resends → Groq extracts items, quantities, deadline, and organization type into structured JSON: each item is matched against a live Airtable price catalogue using normalized units and fuzzy word matching. A priced draft is generated and

LightSpas needed new signups to reliably enter a welcome sequence, stay in sync across tools, and let staff queue social posts from one place. I built this as four linked n8n workflows on Supabase, Brevo, and Buffer. Workflow 1 fires on every Supabase signup. A Postgres trigger using pg_net calls an n8n webhook, which validates the email, upserts the contact to Brevo via a direct API call, and writes the sync status back to Supabase. Signups land in Brevo within seconds, starting a 4-email drip built in Brevo Automation. Workflow 2 mirrors profile edits (name, email) from Supabase to Brevo,

I built a daily opportunity feed built in n8n. Every day at 8:00 (Africa/Lagos), the workflow pulls remote jobs from one source, Remotive, using one search keyword per team member (n8n, backend, React, video editor, copywriter). It cleans the data, removes duplicates and sends each posting to Groq (openai/gpt-oss-20b), which scores fit from 0 to 100 against a five-person team profile. The model also names the best-fit team member and gives a one-line reason. Jobs scoring 70 or higher are sorted, and the top 5 post to Slack as a single message with the link, score, assigned person and reason.

This workflow solves the "DM for order" problem, where a bakery owner reads messy messages and works out items, quantities and dates by hand. Customers can order from a form on the website or by chatting on WhatsApp. The form lets them pick items and sizes from the menu, shows a running total, and opens WhatsApp with the order already written. After my client complained that typing the order format was too much work, the form became the main route. When the message arrives, n8n pulls out the name, phone, items, pickup or delivery, date and time. Groq reads free-text orders and maps them to t