First of all, I want to give a big shout out to the people at LangChain. I have lots of respect to products that empower others, which has always been a designing principle of my own products as well. Let’s next answer the obvious question: what makes Knowlery different to the other knowledge-based question answering chatbots. The answer is that instead of focusing on answering questions, Knowlery is designed to help users to understand domain-specific concepts from many lengthy documents. It focuses on productivity. This means Knowlery’s target audience are students, researchers, professionals who work with many sources of documents to learn and understanding concepts, rather than a customer doing online banking, who wants to get a quick and succinct answer and go. This was inspired by a pain point I faced as a student. I would have collected lots of textbook, tutorials, lecture notes, online articles for a course, but I spent most of time finding the relevant context rather than understanding the concepts. And before you call me nerdy, my life now as a consultant for the big four is exactly the same. So no, being nerdy is not the root cause. This is why the below design elements have been built into Knowlery: 1. A great focus on the contexts extracted, taking up one-third of the screen estate, with most relevant keywords highlighted in a nicely looking blue. I remember Control + F painfully as a student all the time. I love how Knowlery does this for me 2. Knowlery also provides a broader answer from the general knowledge base, and follow up questions and answers, aiming to assist users to better understanding the concepts. For example, if you ask Knowlery “What is augmented reality”, it would also responds with follow up questions such as “benefits of AR”, “How AR is different to VR” etc to further help users with understanding the concepts. In addition, Knowlery also provides a broader and longer answer. This is similar to what ChatGPT does, but only for comparison and further understanding. These features are not implemented by any other products (please comment below if you came across a similar products, in which case I would love to learn from them). And not having these features might be better if the targeted audience may not bother with the why and how, but rather want to get the answer and go. Please let your friends know if you think they can improve their productivity with Knowlery :) Happy coding!

Hunter

📌

3yr ago

PrototypeThis!

Seems useful for teams with big knowledge bases scattered across PDFs. On a design level, the UI button for login with Google is inconsistent with how they are implemented elsewhere. So I think if you fix this, you can get a higher signup rate. It's an interesting tool. Going to try it out later

Awesome, Anthony! Hoping we could support each other today!

Knowlery AI

To explore the how and why rather than a focus on the answer

To explore the how and why rather than a focus on the answer