n8n / automation / ai agents

I build workflows that do the boring stuff, so you don't have to

I'm an automation nerd who spends way too much time thinking about how to save 5 minutes. Over the past year, I've built n8n workflows that have saved me 165+ hours. Here's what I've been working on.

See the workflows ↓ Get in touch
165+
hours saved / year
3
production workflows
4+
AI models integrated
10+
APIs connected

Selected Work

Workflows I've built

01

SMS Expense Tracker

I got tired of logging expenses manually. Now my bank SMS gets parsed by AI and logged to Google Sheets automatically. Takes about 2 seconds per transaction instead of... never getting done.

The problem

Like most people, I never logged my transactions consistently. By the end of the month, I had no idea where my money went.

How it works

Bank SMS → Webhook → Gemini AI parses amount, merchant, type → Routes to correct column → Google Sheets. Groq as fallback when Gemini is being slow.

n8n Gemini Groq Sheets
View on GitHub →
Time saved 30 hrs/yr
Accuracy 98%
Setup time 15 min
Manual input Zero

02

LinkedIn Content Automation

I watch tech videos anyway. This turns those into LinkedIn posts. AI checks if it's relevant to my niche, writes a draft, and asks me on WhatsApp to approve before posting.

The problem

Writing consistent LinkedIn content takes forever. I'd spend 45 minutes on a single post, then get inconsistent for weeks.

How it works

YouTube video → Extract transcript → AI checks relevance → Generates post in my writing style → WhatsApp for approval → Posts to LinkedIn. Smart chunking for long videos.

n8n YouTube API Gemini WhatsApp LinkedIn
View on GitHub →
Time per post ~2 min
vs manual 45 min
Post output 1/day
Setup time 20 min

03

LinkedIn Job Post Outreach

Most jobs are announced in posts before they hit job boards. This scrapes LinkedIn, filters hiring posts with AI, and writes personalized outreach messages. I just approve and send.

The problem

The official job board is a graveyard of old listings. The real opportunities are in posts from hiring managers, but tracking those manually is a full-time job.

How it works

Scheduled scrape every 6 hours → AI classifies hiring vs noise → Extracts contact info → Generates personalized message using my background → WhatsApp approval → Tracks everything in Sheets.

n8n Groq OpenRouter WhatsApp Sheets
View on GitHub →
Time saved 90+ hrs/mo
Leads per day 10-15
AI accuracy 85%
Setup time 25 min

How I think

My approach to automation

I don't believe in automating everything. Some things are better done by hand—the creative work, the relationship building, the thinking. But data entry? Scheduling? Scraping? Those are a waste of human attention.

// My automation philosophy
if (task.isRepetitive && task.frequency > "weekly") {
  return automate(task);
}

// Always keep humans in the loop for decisions
workflow.addStep("whatsapp_approval");

Every workflow I build has a human approval step. I don't trust AI to make decisions for me—but I do trust it to do the prep work so I can make better decisions faster.

A bit about me

I'm a developer who fell into automation after realizing I was spending more time on busywork than actual work. Now I build n8n workflows that handle the boring stuff so I can focus on things that matter.

Tools I use

  • n8n (self-hosted)
  • Gemini, Groq, OpenRouter
  • Google Sheets API
  • LinkedIn API
  • WhatsApp Business

What I care about

  • Reliability over cleverness
  • Human-in-the-loop design
  • Actually useful automation
  • Clean documentation
  • Open source

Want to work together?

I'm open to freelance automation projects or just chatting about n8n.

goutamaswani43@gmail.com LinkedIn GitHub