Business mindset in Aurora Projects

In the direction of targeting more small and medium enterprises (SME), we wanted to provide no-code (or low-code) NLP tools to reach SMEs businesses and customers with in-house production.

Our business mindset of forming a new product in Aurora is based on lean startup principles. In the very first step, our AI team works with the delivery team to understand the unique problems for a specific group of customers and based on that to form our vision of the product. In this step, we need to balance between our viable technologies and resources and the features that will be compelling for our customers.

In the case of our Aurora product, after working with the delivery team, we found out that it is really hard to develop one generic solution that can directly serve all clients as each client has a different need when it comes to applying NLP in their business process. Additionally, we cannot approach SMEs with POC projects tailored to their needs since their budget is really limited. In order to reach these clients, we proposed to develop an Information Extraction platform to help SMEs developing their own AI solutions. Our platform enables the clients to train their own AI models from their labeled data, choosing the models that fit their businesses for easy integration with their systems.

One of the difficulties approaching the business mindset in this project is the lack of feedback from our clients in the early development phase. This limits the understanding of clients' needs to propose a suitable product. To address this we worked with our delivery team and came up with minimum viable features for the early phase. This helps us to bring the product to our clients faster and avoid developing the features they don't need. After finishing our minimum viable product (MVP), we reached out to our early adopters to survey their interests as well as using their feedback as insight to continuously improve, and propose more features.

Another challenge when applying a business mindset to this project is how to find the balance between our current research direction, current projects, and the new product as the human resource in an AI startup is really limited. To overcome this, we design our MVP to utilize what we already developed in our current projects as well as choosing the MVP features to not only bring value to our client but also provide tools to advance our research.

On the final note, based on our experience, a business mindset is essential for us when it comes to the development of our new products. We have to not only put ourselves in our client's shoes but also strike the balance between our resources, priorities, and what clients want.

A lean startup process, where we pick a minimum viable product to continuously build and improve based on our clients’ feedbacks

Matthew

AI Researcher - Aurora Team