SecondResponse

💡 Inspiration

A large portion of the response on the current questionnaire is invaluable due to vagueness and lack of interpretability. We want to improve the current way of doing questionnaires at scale without sacrificing their quality.

💥 What it does

Our project opens the doors for more thorough data collection through context-based follow-up questions.

🛠️ How we built it

Quick proof-of-concept done through streamlit Frontend built by JS, HTML and CSS Backend built by Flask utilize openAI API

🧗‍♀️ Challenges we ran into

Unfamiliarity with the integration of front-end and back-end development Prompt engineering is a novel task to navigate

🏆 Accomplishments that we're proud of

Achieve high-quality, sensible interview questions generated by chatGPT Successfully learned and integrated Flask library (in 36 hours) Practical, well thought idea, and decision to have one polished feature

📚 What we learned

Through close and intense development with AI tools, we learned the vulnerability and malleability of AI and the ethical problems from it. A diverse skillset (tech stack) is necessary for practical software development

💭 What's next for secondResponse

Extract common themes from survey results using the AI model Enable export of survey results in standardized form Create a database where results are stored and can be accessed by the surveyor

Built With

• css
• flask
• html
• javascript
• openai
• python
• streamlit