Gemini API
The Gemini API is Google's AI service, and I use it to build chat assistants and text processing into websites and apps.
How I use Gemini API
The assistant answers visitors based on the company's own information and can also write or summarise text.
Visitors get an answer straight away, at night and at weekends too, without waiting for an email reply. The business doesn't have to answer the same questions by hand every time, and the assistant points people to the right page, service or contact.
In practice I give the assistant instructions and verified company information, so it stays on topic and speaks in a tone that fits the brand. On this portfolio, the chat assistant answers visitors' questions about my projects, skills and services.
Because AI can make mistakes, I set clear topic limits and give the assistant checked facts instead of letting it rely on general knowledge. I build the Gemini API into React apps, and the information about projects and services can come from a database such as Supabase.
Used in: the chat assistant on this portfolio.
Frequently asked questions about Gemini API
What is the Gemini API?
The Gemini API is Google's service that lets an app send text to a Gemini AI model and get a response back. It's used for chat assistants, writing and summarising text, translation and document processing.
How do I add an AI chat assistant to my website?
You add a chat window to the site that sends visitors' questions to an AI model together with instructions and information about your business. That way the assistant talks about your services, opening hours or products instead of anything at all.
Can you add an AI assistant to my website?
Yes, a chat assistant running on the Gemini API is live on this portfolio, answering visitors' questions about my projects and services. I can build the same kind of assistant into a company's site or app, using that company's information.