An AI Assistant That Gives Gas Station Teams Clear Troubleshooting Answers
A RAG-powered chatbot built into the mobile app of a US gas station community
About the customer
Our client is an online community of gas stations across the United States, connecting vendors, service providers, industry experts and online stores. Its goal is to help gas station owners work more efficiently, simplify processes and maximize profits. We built an AI chatbot using OpenAI’s large language models (LLMs) to answer user questions about the standard documents for operating and troubleshooting gas station equipment.
What needed solving
Language barriers
A significant portion of the US gas station workforce comes from non-English-speaking countries and has limited education
Hard-to-follow guides
Standard procedures and troubleshooting guides were hard for them to understand and follow
Answers on the job
Workers needed accurate, easy-to-understand information quickly, during their daily tasks
How we built it
Answers grounded in documents
A RAG (retrieval-augmented generation) pipeline gives the LLM the relevant context from the documents, so answers are accurate and concise.
Knowledge base
About 100 documents, including user manuals, troubleshooting guides and registration manuals.
AI models and tools
OpenAI GPT models, with LangChain and LlamaIndex connecting the models to the documents stored in a vector database.
Application and hosting
Python and Flask power the application’s API, hosted on AWS (Amazon EC2), with PostgreSQL and MongoDB storing the data.
Mobile app integration
The chatbot is built into the client’s mobile app.
Clear, concise answers for gas station owners and employees
Technologies
Domains
Have a Query? Let's Have a Conversation
Do you have a project in mind? We’d love to know what it is. Email us or send a message from our contact page.
Have a similar problem?
Tell us what you are trying to solve and our team will get back to you.