Inspiration
On average, product managers (PMs) roughly spend around 7 hours a week on time consuming tasks that can be easily completed through the use of AI-driven applications. This inspired us to design a web application called ProSolve to improve the overall productivity of a product manager by automating simple, time consuming tasks.
What it does
ProSolve is an AI-powered task planning web application that improves PMs overall productivity through automating tasks like performing risk analysis, identifying user/customer needs, and developing possible recommendations/critiques. Our web application does by making calls to our LLM Groq, which processes information from user inputs to generate actionable insights, prioritized tasks, and strategic guidance.
How we built it
For the front-end architecture of our project, we used HTML, CSS, and JavaScript to add text, images, buttons, styles, colors, functionalities, and user-interactivity to our web application. For the back-end architecture of our project, we primarily used python to handle server-side logic and process user requests while using SQLite and SQLModel to manage storage of past tasks of a PM. To develop AI-driven responses, we implemented the Groq API key which allowed for llama-3.3-70b-versatile to process user-input to effectively automate tasks.
Challenges we ran into
A major challenge we ran into during the development was effectively making sure Groq API calls in the front-end was effectively received on our back-end of the project. During this point of the contest, user-input on the front end of our web application failed to initiate any calls to the Groq API which gave us a lot of trouble because this resulted in no response from llama-3.3-70b-versatile when a user provided a task and details of the task.
Accomplishments that we're proud of
A major accomplishment we are proud of is being able to program a highly impactful project in such a budgeted time setting, which is a new feat for both of us. Another accomplishment we are proud of is being able to effectively learn more about full-stack development through the process of creating this web application. Since we are both seeking careers in tech in the future, we highly value these skills and treat them as great achievements.
What we learned
During this hackathon we learned numerous things about full-stack development and being able to build back up after large mistakes. We learned a great amount of new concepts like SQLite and SQLModel and how to use them to make effective storage mechanisms in your project. This allows for us to provide PMs that would use our web application to have access to prior tasks and dates in which these tasks have been completed. Aside from the technical field of software, we learned about coming up from large mistakes. For example, in the first 4 hours of our project, we designed our back end with a lot of errors meaning we would have to restart from scratch. Instead complaining or being demotivated by this major setback, the lack of time left to create a working product motivated us to develop a more robust front-end framework that would serve as a major improvement from what we had originally.
What's next for ProSolve
A major next move for ProSolve would be to add more capabilities of our web application so we can effectively automate more time consuming tasks that may burden PMs and their overall effectiveness.
Log in or sign up for Devpost to join the conversation.