AI Search Application for Researchers

Background

Managing unstructured data is a massive pain point across just about any industry, and for just about every individual. People experience “information overload” online and have messy file folders on local drives. Corporations and universities have massive repositories of content just sitting there.

The company I was working for had developed a (linguistics rules-based NLP) search engine that would read, interpret, and learn from written text. It could be pointed to the Internet or any text-based repository to generate knowledge on demand, helping bring order to all the chaos in information in the digital space.

The AI learns from all the information it reads, and then stores it in the form of a massive knowledge network consisting of the concepts its encountered and how they relate to one another. The AI and knowledge network work like this:

The AI recognizes the main concepts and supporting ideas from each document.

The AI had an inventory of cognitively significant relationships that it could identify between concepts by using linguistic analysis. It would generate a knowledge network of this information, which was both machine and human readable.

The AI connects information from different sources across texts using this network.

The network contains knowledge and can be queried.

The AI learns from everything it reads, continuously building a bigger and smarter network.

Vision and Approach

We were building an application that would enhance people’s experience engaging with information in digital spaces. The user journey would be simple because the AI would do the bulk of the work. Users would input their search into the app and within moments receive a personalized knowledge collection, which would provide the following:

The AI builds a Table of Contents to address each search, and organizes relevant information accordingly.
The AI builds a knowledge network from the relevant information, which users could explore through an interface.
All information would be in a useful and usable format, and easily traceable back to its source.

My Role as Product Manager

Being the first employee hired to work on the application marked my experiences. We were a startup, meaning we were building from the ground up and it was all hands on deck. My primary role was to build out the B2C search experience around the AI technology. I would also partake in the B2B end of our business and product development, where the work focused heavily on dealmaking: use case and product discovery, POCs, and pilots.

In my day-to-day role:

I conducted research and surveys to establish a strong understand about the business problems we were addressing, our users and their pain points.
I worked closely with the core engine team to develop a deep understanding of the AI system, and stayed current with engine advancements and new capabilities.
I recognized opportunities where we could apply our AI to solve user pain points and elevate user workflows.
I demonstrated our AI’s capabilities by composing demos.
I sought and constantly incorporated feedback from users and stakeholders.
I coordinated the work that was required with leads across engineering, linguistics, design, and marketing using methodologies that would be best described as Agile.
I managed the collaboration between internal and external stakeholders.

The Application

Sample input query:

The impact that Covid-19 had on mental health, and how different countries addressed the public health crisis.

Sample output: