Build an AI Search API with LangChain, FastAPI, and MariaDB

Description

Alejandro Duarte shows how to use LangChain's MariaDB integration (langchain-mariadb) to implement vector search. This demo compares classic text search vs. semantic search using Gemini for evaluation. ►Github Repo: https://github.com/mariadb-developers/langchain-fastapi-mariadb-webinar-demo ►Try MariaDB cloud: https://mariadb.com/cloud-get-started/ ►Book: https://mariadbfordevelopers.com Chapters: 0:00 Setting up a free MariaDB serverless instance 0:54 Connecting to MariaDB from VS Code 1:32 Creating the database and importing data 3:07 Creating a Python project 3:53 Connecting to MariaDB from Python 5:07 Implementing text search (fulltext index) 7:24 Running the application 8:00 Creating a VectorStore 10:54 Ingesting vector embeddings 12:20 How LangChain and MariaDB store vectors 13:31 Implementing vector search (semantic) 14:59 Comparing results ►Explore MariaDB Products: https://mariadb.com/products ►Check out the MariaDB DeveloperHub: https://mariadb.com/developers/ ►Get in touch with MariaDB experts via Slack: https://mariadb-community.slack.com/ (invite: https://r.mariadb.com/join-community-slack) MariaDB is making a big impact on the world. Whether you’re checking your bank account, buying a coffee, shopping online, making a phone call, listening to music, taking out a loan, or ordering takeout–MariaDB is the backbone of applications used every day. Companies small and large, including 75% of the Fortune 500 run MariaDB, touching the lives of billions of people. With massive reach through Linux distributions, enterprise deployments, and public clouds, MariaDB is uniquely positioned as the leading database for modern application development. #MariaDB #Database #AI #Data #SQL #Cloud #python #langchain #gemini #programming