SenseAI

• 

  Product Logo

Inspiration

We have had family members who worked in call centers and have struggled with difficult customers and an overwhelming workload. We wanted to help these employees like them manage stressful conversations while maintaining professionalism and efficiency . Seeing AI improve industries, we wanted to build something to use it to compile and analyze data at speeds far beyond human capability, and apply it in a way that would make life easier for sales representatives and customer service employees.

What it does

Our platform is a B2B SaaS product for customer service. Acting as a one stop shop for customer service, it analyzes calls both in real time and retrospectively, compiles insights from existing company data, and provides employees with actionable feedback to improve performance and customer satisfaction. Additionally, it provides complete business statistics and suggestions for users. For additional guidance, there is an AI chatbot that works at any time, including during a call, to answer any questions that an employee may have, whether it be about their performance or specific questions about company policies.

How we built it

We started by planning out the features we wanted to include and the code flow. We then created a basic UI using ReactJS and started developing the backend piece by piece with expressJS to make each feature come to life. We gave it some dummy data from a publicly available spreadsheet from kaggle to base the business analytics off of. We integrated it with the Gemini and OpenAI APIs to provide responses to the users.

Challenges we ran into

Getting our AI chatbot to generate meaningful and context-specific responses was difficult. It often repeated outputs at first. We also faced challenges while training our call-analysis model to understand tone, sentiment, and customer intent accurately.

Accomplishments that we're proud of

We achieved real-time integration of our model for data analytics and successfully implemented live call analysis within the short timeframe. It was rewarding to see our system provide immediate, helpful feedback during test calls.

What we learned

We learned how to connect multiple technologies including the frontend, backend, and AI into a cohesive workflow. We also gained hands-on experience with real-time data handling and integrating large language models for business-driven use cases.

What's next for SenseAI

We plan to add more features, for example, generating employee performance reports for managers, improving model accuracy through fine-tuning, and expanding data visualization in the dashboard. We also plan to have this work as a personalized platform for each user company to tailor to their needs and policies. This would allow the AI chatbot to give specific suggestions based on the company and what it offers. There will also be a feature where managers can view reports on the performance of each of their employees in order to more quickly improve the performance of their team and respond to problems.

Built With

• gemini
• node.js
• openai
• react
• vader