Leveraging Generative AI for Seamless Conversations with PDFs
I've founded, designed and launched PDF AI from 0-1 from research to production-ready.
A side-project to leverage generative AI to interact with documents like never before, speaking with them.
Role
Founder / Product Designer
Duration
Present
Team
Edward
Software Engineer
Context
PDF AI enables you to Chat with any PDF document.
From legal agreements to financial reports, PDF AI brings your documents to life. You can ask questions, get summaries, find information, and more.
Problem
PDFs, they are EVERYWHERE. Interacting with PDF documents remains a significant challenge for many people like students, lawyers and professionals. PDFs, being static and non-interactive, often create barriers to information retrieval, understanding, and communication.
The core problem lies in the lack of dynamic interaction between the user and the content within these documents. Users frequently need to extract, interpret, and inquire about information in PDFs, but the traditional methods of engaging with these documents are cumbersome and time-consuming.
This challenge becomes more pronounced in environments where quick information retrieval and decision-making are crucial, such as in legal, academic, and corporate settings. Users often find themselves manually searching through lengthy documents, struggling to locate specific information, or having to use multiple tools to interact with various data points within a PDF. This leads to inefficiencies, a higher likelihood of errors, and a generally frustrating user experience.
Goal
Enable users to have a conversation with their documents for better learning experience
Success Metrics
Growth in MRR and ARR is a strong indicator of a healthy SaaS business.
1
Monthly recourring revenue (MRR)
tracking the percentage of leads that convert into paying customers. It helps measure the effectiveness of the sales and marketing funnel.
2
Lead converion rate
The percentage of customers who cancel or do not renew their subscriptions during a given period
3
Churn rate
Description
4
Higher CLTV indicates a more valuable customer base.
Solution
Introducing PDF AI, a tool that enables you to chat with any PDF document, ask questions, retrieve info and even have the option to chat to multiple PDFs.
1
Ask questions
2
Cite sources
3
Summarise findings
4
Chrome extension to talk to PDFs everywhere
Design decision: Tooltip
After the MVP launch, we noticed users would highlight the PDF content often, this led me to speak with 5 users that explained they'd like an easier way to get more accurate answers without having to type overly specific questions so we introduced the ability to highlight portions of the text and click "Explain" or "summarise" which received wide-spread adoption quickly.
Design decision: Chrome extension
Further research led us to learn that users would go to Google Drive, search for the PDF on Google, download it and upload it to the platform. We figured the best way to streamline the experience further was to design and build a chrome extension that is accessible whenever a user opens a PDF online.
Outcomes
Currently 12% every month, with most of our traffic coming from SEO.
1
Growing user base
Revenue is steadily growing as we've introduced new features such as the chrome extension, the ability to talk to multiple PDF's
2
Growing revenue
With the the amount of competition, users churn to try other options and find the right product for their use case
4
18% Churn
Takeaways
Before launching the MVP, I should've spent more time on the distribution channels since we launched and it took a while for SEO to kick in! Overall, It was great fun!