About Us

Introduction

My name is Shashank Agarwal, and I used to run a company called Pipfeed, a news-reading mobile app that utilized user behavior to recommend articles. Unfortunately, due to high customer acquisition costs, the app didn't succeed, and I had to shut it down. However, I realized that some of the services I had created could be valuable to others. So, I deployed these services to AWS, resulting in a URL parsing API.

Early Struggles and Progress

In the initial six months, the earnings were meager, just around 10 dollars per month. But gradually, the API gained traction, and I am now making over $1500 per month. To monetize the API, I decided to use RapidAPI, a marketplace for various APIs, even though it takes a 20% commission, leaving me with $1200 from the $1500 earnings. Nonetheless, the profit margin remains at approximately 90% on the $1200, with the main cost being RapidAPI's commission.

Creating the Metering Platform

To reduce this cost, I decided to develop a simple metering platform for my API. However, the process proved challenging, with many components needing scaling. Recognizing this as a substantial project that would require significant time investment, I thought about offering this platform to other API providers. To test this idea, I asked Arjun to reach out to other API providers on RapidAPI, and our intuition proved right.

Promising Demand and Customer Selection

More than 25 API providers joined the waiting list by filling out a form detailing their average monthly API calls, total users, etc. From this list, we selected ten customers with substantial API call volumes. After calculating their monthly earnings using their public pricing plans, we were surprised to find that these ten customers collectively earned over $50,000 per month, using only the cheapest plan.

Scalability and Revenue Projections

With this promising demand, we determined how much money we could make and how to go about it. To ensure low server costs, we designed the system to scale easily. The total API calls for these ten customers amounted to over a billion API calls per month.

MagicAPI's servers were load-tested at 13K TPS, capable of handling at least 36 billion API calls at peak loads.

Overall, it's an exciting prospect, and I'm looking forward to what the future holds for MagicAPI.

Last updated