Andrew Ng, the godfather of deep learning has come up with a new course called ‘Building Your Own Database Agent’, this time, in collaboration with generative AI giant Microsoft.
Offered for free, this beginner-level course, teaches you how to interact with tabular data and SQL databases using natural language, making data analysis more efficient and accessible.
Led by Adrian Gonzalez Sanchez, data and AI specialist at Microsoft, in one hour, you will gain hands-on experience with the Azure OpenAI Service, learning techniques such as RAG and function calling, among others.
Under the Hood
While you are pursuing the course, you will develop an AI agent capable of using natural language to query and extract insights from databases.
The course content includes learning how to customise knowledge levels with Azure OpenAI Service to build your first AI agent. You will deploy an Azure OpenAI Service instance, test the API, and use LangChain to set up an orchestration engine that enables various scenarios. Additionally, you will learn to load and analyse tabular data from CSV files using natural language queries and apply this knowledge to analyse your own CSV files.
One of the key components of the course is implementing LangChain agents to connect to a provided SQL database and building a database agent that translates natural language into SQL code. You will also use Azure OpenAI Service’s function calling feature to send queries to databases, improving the efficiency and security of your SQL agent.
The course further includes working with the Assistants API and testing it with function calling and code interpreter features, enabling you to connect to SQL databases and create your own database agents more efficiently.
By the end of the course, you will be equipped with both the technical knowledge and practical experience necessary to implement similar systems in your projects or organisation, enhancing the efficiency and accessibility of data interaction and analysis.
This course is tailored for developers, data professionals, business analysts, and practically anyone interested in improving their interaction with databases without needing advanced SQL queries.
While it is recommended to have familiarity with Python programming and databases (CSV files and SQL), it is not a prerequisite.
This is not the first time that DeepLearning has collaborated with Microsoft to provide free learning content.
For example, “How Business Thinkers Can Start Building AI Plugins With Semantic Kernel,” guides you in using Microsoft’s open-source orchestrator, developing business applications with LLMs, and leveraging tools like memories and chains. Another course with Microsoft called “AI Agentic Design Patterns with AutoGen” covers building multi-agent systems for complex AI applications using the AutoGen framework.
