TechShopper.AI

Tech Shopper AI is your tech-savvy family member. It uses what you know to tell you what you need, all at the best price! Available to most through a web app, and gives accurate results every time

Comment

Inspiration

Our biggest motivation for this project is because we wanted our own family and friends pay for quality tech without breaking the bank. We have found that many lack the time, knowledge, and resources to research what tech would work for them and their budgets. So, we decided to bridge that gap to improve others lives while saving their money with a new friend named TechShopper!

What it does

TechShopper combines OpenAI with an efficient UI to recommend the best purchase for the customer. It considers the user's current tech and their preferences to recommend a product that is compatible and is the best option for that specific user. Once a recommendation is made, TechAI does the heavy lifting and finds the best price from reputable retailers across the web.

Our Team

TechShopper was made by a group of 3 people. Utilizing Python, Jake utilized Microsoft Semantic Kernel in order to easily access the LLM of our choice, GPT4. Alexandra produced the user interface by implementing a blend of Django and Bootstrap, while Nicolas took user requests to scrape price data from the web. User preference data is stored in MongoDB Atlas, to improve their shopping experience if and when they return in the future.

Our Challenges

After hearing RBC's challenge we decided to completely change our original project idea, step out of our comfort zone, and dive in head-first. All three of us had little expertise in coding with LLMs, so our project functioning is an accomplishment in itself. Consistently, we struggled with managing dependencies, package differences, gaps in knowledge, and device usability. However, with teamwork, communication, and many revisions to our program, we pushed through.

What we gained

A huge lesson we learned, was our ability to work in a group, especially when we all had limited knowledge in the algorithms, databases, and frameworks we were going to work with. This project has given us the opportunity to improve our skills in AI development, object-oriented programming, and database management.

What's next for TechShopper.AI

The next steps for TechShopper is optimizing our model, outputting better recommendations, and implementing accessibility through TTS, keyboard shortcuts, so customers of all ages and abilities can use our model.

Built With

Share this project:

Updates