How I Built My First AI Workflow Without Writing Code
Finally I am going to share my workflow with you guys!
Finally I am going to share my workflow with you guys!
Now that we’ve covered the foundational tools — APIs, tokens, and cost management — it’s time to dive into what really makes things exciting: no-code automation workflows. If you’re a biomedical scientist like me, with zero formal training in programming, the good news is that you don’t need to become a software engineer to automate your research workflows. You just need the right tools and a bit of curiosity.
Today, I’m sharing my first experience with building an automation pipeline using no-code platforms, and how I gradually transitioned from Make to n8n for building more advanced workflows.
Popular No-Code Automation Tools
There are several well-known services that offer no-code automation capabilities:
- IFTTT (If This Then That)
- Microsoft Power Automate
- Apify — great for web scraping (I use this in my own workflow!)
- Make (formerly Integromat)
- Zapier
- n8n
Since I’ve only used the last three tools, I’ll focus on comparing them to help you pick the right one based on your technical comfort level and goals.
As you can see, for a biomedical researcher with little to no coding experience, Make offers a great balance between ease of use and functionality. n8n, on the other hand, offers significantly more flexibility and control but comes with a steeper learning curve. That’s why I started with Make and only moved to n8n after gaining more experience and confidence.
Building My First Workflow: A 6-Step Tweeting Bot in Make
Instead of diving straight into building a complex literature analysis pipeline (which felt intimidating), I followed a popular online tutorial to create a simple but functional 6-step Twitter bot using Make. This helped me understand the fundamentals without getting overwhelmed.
Here’s what my Make automation does:
- Reads a list of cryptocurrency news links from a Google Sheet
- Uses HTTP requests to retrieve article content
- Sends the content to ChatGPT 4o-mini for summarization
- Passes the summary again to ChatGPT 4o-mini to write a tweet
- Formats everything into JSON
- Posts the tweet on X (Twitter) and sends a Telegram notification, while updating the status in Google Sheets
Visual Simplicity: Why Make is Beginner-Friendly
Make’s visual interface is highly intuitive. Each task (called a “node”) is represented by a circle. In my case, I used 9 nodes total, including one router node that requires no configuration.
Each node has a simple setup interface. For example, when using HTTP to fetch article content via Jina.ai, I just followed the step-by-step guide from their website and plugged in the required parameters. No coding. No confusion.
Scheduling & Optimization
After setting up and testing the workflow, I configured Make to run this automation every 6 hours. Here’s why:
- I didn’t want to spam my X followers with too many tweets
- Make’s free plan offers 1,000 operations/month. With my 8-node setup running 4 times per day, that’s 32 operations daily or 960/month — just under the free plan limit. If I scheduled it to run 5 times a day, I’d exceed the quota.
This is a great example of how even resource constraints can teach you how to optimize workflow design!
Free Template for You to Try
To help you get started, I’m sharing the template for this Twitter bot built in Make . If you’d like to try building your first automation without code, this is a fantastic way to learn the ropes. And don’t worry — if you run into setup issues, I’m happy to help. Just reach out!
In upcoming articles, I’ll walk you through how I transitioned from Make to n8n, and how I began building automated research tools for literature review, scientific writing, and grant drafting.
Follow me to learn how automation and AI are transforming the way we do biomedical science — one no-code workflow at a time!
