Customer Service Sentinel

Inspired by the project I contributed to during my internship, my team and I developed an advanced chatbot assistant for our senior project, which we named Customer Service Sentinel (CSS). The project was driven by the goal of creating a customizable chatbot platform that companies could easily tailor to their specific needs, even if they lacked technical expertise. We aimed to simplify the process of setting up an AI-powered customer service solution by enabling users to upload custom assistant photos, configure data, personalize color schemes, and seamlessly integrate the chatbot into their social media platforms or company websites.

As the team leader, I was responsible for overseeing the project from conception to deployment, ensuring that all aspects of development ran smoothly. Although I contributed to various components, my primary focus was on the back-end development and the chatbot’s core functionalities. The front-end was designed and built using Next.js and TailwindCSS, which provided a modern, responsive, and user-friendly interface. On the back end, I used Flask to create the API responsible for managing and delivering chatbot services respectively. I also used Docker to deploy the finished project.

The chatbot itself was powered by OpenAI’s API, and we integrated it with Langchain to significantly boost its capabilities. By using Langchain, the chatbot could embed and utilize external knowledge, making it flexible and adaptable for a variety of industries. The chatbot could be trained using specific company data, allowing it to deliver tailored customer service interactions. Additionally, the chatbot was designed for easy embedding on company websites through simple code snippets or integration into social media platforms, offering companies flexible deployment options.

The Customer Service Sentinel simplifies the adoption of AI-driven customer support solutions while empowering businesses to customize their chatbot service without needing deep technical knowledge. This project was the culmination of our team’s efforts to develop practical tech solutions that bridge the gap between innovation and accessibility.

We tested the chatbot on Night Login’s website, where it operated under the name NL Bot. Unfortunately, due to a lack of funding, we were unable to maintain the service beyond the timeline of our senior project. However, the project remains available for exploration, and you can find more details on our GitHub page.