In 2017, Andrei and I began to build our vision of a comprehensive platform that powers conversational experiences based on his time building American Express’s conversational AI platform from scratch. Our first step in the journey was creating designs and figuring out what we wanted the platform to do. We began setting up systems with myself working on the backend infrastructure. Our other co-founder Peter joined us in fall of 2017 and began helping us with turning our early designs into a working web application.
Since it’s conception, the platform was always meant to be usable by non-technical people no matter the final output – be it an Alexa Skill, smart IVR or chatbot. Building tech that is easy to use helped us differentiate ourselves in the market as the other companies continued to build challenging platforms that are developer focused. I would work on building the NLX platform during evenings and weekends during the first few years because I still had a full-time job.
From the beginning, we crafted the platform to be secure, scalable and cost efficient by taking advantage of managed services from Amazon Web Services (AWS). We make use of over 50 AWS services with the idea of using the right tool for the right job. This decision has enabled us to build a lot of resilient technology very quickly, which gave us and our customers a lot of confidence. For example, if a large company like American Express or Chase wanted to use our platform, they could run their workloads at full capacity due to how we architected our systems. Unlike other companies, NLX did not start with a prototype. We built it like we would any production system. We made sure to do things with 100% effort now so we would not regret it later. Most of the code that was written three years ago is still relevant, despite the rapid advance of technology.