By his own count, Irzan Raditya had failed at four startups before he built the one that lasted. He told KrASIA he treats each failure as a lesson: either you win or you learn.
The earlier attempts started around 2012, when he was living in Germany and building companies as a software engineer. He came back to Indonesia and in 2015 launched YesBoss, a virtual assistant you could text, a concierge in your pocket. It was a charming idea and, as it turned out, a hard business.
The company he built with Wahyu Wrehasnaya and Reynir Fauzan was reborn as Kata.ai, a platform for building chatbots that understand Indonesian.
What the YesBoss story hides is the cost of the next step. A consumer assistant needs humans behind the screen, and the numbers did not hold. Irzan decided to leave consumers behind and sell to businesses. Pivoting meant changing the product and the people: some employees whose jobs no longer fit had to be laid off. He has described that period as part of learning to put people first, product second, then profit.
The company he built with Wahyu Wrehasnaya and Reynir Fauzan was reborn as Kata.ai, a platform for building chatbots that understand Indonesian. Customers were large organisations, banks, telecoms, retailers, that needed to answer millions of messages. In 2017 it raised US$3.5 million in a Series A led by Trans-Pacific Technology Fund, with Red Sails Investment and the angel Eddy Chan taking part.
The next years were quieter. Language models were still hard in Bahasa Indonesia, and AI talent was scarce. A Series B arrived in November 2020, led by TPTF and MDI Ventures with Buana Investama, though the amount was never disclosed. Kata.ai used it to push into commerce, healthcare and insurance and to launch a tool for small businesses to automate chat sales.
Irzan has called himself a person in permanent beta, a phrase that sounds casual and describes the whole arc: five attempts, one pivot, layoffs, and a product still evolving.
For a first-time founder the takeaway is plain. The company that lasted was not the first idea. It was the fifth, and it only appeared after he was willing to shrink the thing he loved.