To build our software, we started doing small-scale customer surveys to better understand customer needs and how our solution could address them. Our pitch to potential customers was generally met with enthusiasm and excitement. Subsequently, we started presenting our vision to potential investors. Investors, in turn, offered two responses: On the one hand, they recognized that writing practitioners showed genuine interest; as writers, they also could relate with advantages of the Lexprompt solution. On the other hand, investors as well as certain experts and software developers voiced skepticism about the technical feasibility of the proposed solution. Some thought a powerful server would be required to produce useful results.

We considered a different approach. To convince the technical skeptics, we proceeded with the development of prototype software that would test key aspects of the solution. We also hoped to give users a realistic feel of what it would be like to use the commercial software.

The prototype software exceeded all expectations. The prototype was built as standalone software that included a simplified word processor and a suggestion window that displayed relevant text from existing documents. The prototype used a custom database and natural language processing (NLP) algorithms, including retrieval algorithms that accounted for synonyms and word morphology. The software also used NLP algorithms to resolve existing documents into text, which populated the custom database. Then, as a user typed in the word processor window, the prototype displayed relevant text suggestions in real time.

We ran various tests on the software and the results were impressive — the responsiveness of the prototype was better than anyone had hoped for. Additionally, the prototype was effortless to use and promptly displayed suggested text.

After testing the prototype software, we knew that the technical skeptics were wrong. More importantly, we knew that users would enjoy using Lexprompt software. As the prototype proved that the Lexprompt solution could be a reality, we entered the next phase of our plan.