Atexto is the machine learning data and training services platform that looks to improve the accuracy of automatic speech recognition systems. They make it possible for companies to automatically understand customers’ and users’ voices (no matter their accent) over the phone or Zoom.
Reaching out to Contxto before anyone else, the startup announced the opening of its Enterprise Portal to the entire Latin American community. It is aimed at those interested in machine learning and data science, with a focus on natural language processing and automatic speech recognition.
That is data scientists, academics, linguists, students, software engineers… And the launch is free.
So, I got on a call with Atexto in order to fully understand how it is that a free product can lead to the sort of ambitious growth the company is looking for.
Series A with a free product
Co-Founder and COO of Atexto, Esteban Gorupicz (pictured), is very clear about the company’s medium-term goals:
I’m sure he means it in terms of seduction, but such a conquest still depends on two crucial factors.
The first is capital. Indeed, Gorupicz revealed to Contxto that Atexto is on track to raise its Series A round in 2021. This comes on the heels of a US$700,000 seed round in January, which means the company has received a total of around US$1 million so far.
The second condition for Atexto’s conquest concerns the company’s ability to get a foothold in the US market. Gorupicz thinks they’re not quite ready to take the leap just yet.
However, expand they must. The American market holds riches that are simply not yet available in Latin America. Moreover, the idea is to raise the Series A round in the US in order to scale successfully.
It was at this point that I worried out loud to Gorupicz: Isn’t it a bit sad for Latam to see so many successful startups scale up and out of the ecosystem?
Atexto’s COO assured me that their plan was precisely to do the opposite: The idea behind launching a free (rather academic) will both keep the company’s roots in the region and train this bot trainer for its incoming expansion.
Atexto riding the adoption wave to Series A and B-eyond
The plan is to tether the startup’s growth and funding model to what Gorupicz has identified as a clearly emerging process of machine learning adoption.
|Machine learning adoption stage
|Free and open Enterprise Portal
|Series A (2021)
|APIs for mass market
|Series B (2022)
The first column refers to how advanced adoption is at any one stage. The creators refer to big specialized companies. The fine-tuners and trainers are those that adapt existing software for their own uses. And the final stage of adoption allows for anyone to plug and play with machine learning tools.
The second and third columns refer to Atexto’s past, present, and future visions in terms of solutions and funding. Not how the company has aligned to the cycles of adoption that they hope to not only ride, but to potentiate in Latin America.
A “quid pro quo” pro
The ultimate idea is that, by the time Atexto gets ahold of a Series B, it will be selling voice AI APIs to individuals and small businesses. And thus, the path to broad market adoption will be complete.
By launching its latest free product in Latam, the company is potentiating this adoption cycle. On one hand, it gives cutting-edge voice analysis machine learning tools to anyone who wants to tinker with them. And, on the other hand, Atexto will be beta testing its wares on a knowledgeable demographic of local experts.