AI-Powered Programming With Google Jules AI Coding Agent

Learn how AI-powered coding agents are non-deterministic, producing varied results even with the same prompts. Practice reviewing, testing, and guiding AI to support your decisions while staying in charge.

Learn to use Jules in the browser without installation, connect GitHub, create a repository, and generate a recipe app by drafting and refining a plan, then review the code proposal.

Explore prototyping with a React frontend and Python Flask backend using Jules AI, run locally, debug CORS issues, and learn how AI-generated code may need manual corrections.

Learn how to craft clear, specific prompts for AI coding agents like Jules, providing enough detail and examples to guide plans, specs, and tasks while avoiding ambiguity.

Explore how to use Jules to solve common software engineering problems across languages. Understand that the course will not delve into non-deterministic generated code and will focus on Jules's capabilities.

Build a prototype application using Go and GraphQL with PostgreSQL, collaborating with Jules to implement GraphQL operations for loan submissions; the app will work but not production-ready.

Prototype a Go and GraphQL loan application by creating a non-blank repository with Go module, skeleton main, GraphQL resolvers and servers, unit tests, and a readme with sample requests.

Identify and resolve issues in generated code, from compilation errors to mismatched GraphQL schema fields, such as missing customer id, and learn to discard flawed code and branches.

try a different approach: generate a GraphQL schema from the spec with ChatGPT, feed it to Jules to build the server, then fix null pointer errors and publish branches.

Run the GraphQL health check and manage a loan application via Postman. Verify submit and cancel operations and assess Jules AI agent’s speed and feedback within an hour.

Use PostgreSQL as the data storage, separate tables with cardinality, and generate SQL to support the Go application's GraphQL schema, including migration files.

Remove unused imports, implement a conditional enum type check in PostgreSQL, fix queries to loan_status and date_of_birth, adjust the PostgreSQL connection string, and avoid soft deletes on cancellation via GraphQL.

Review how Jules helps create an application from scratch and offers alternatives when it cannot solve a problem. Give feedback to guide fixes, review the directory structure, and manage bugs.

Explore a Python repository with undocumented or poorly documented files, identify rest api and data analytics components, and collaborate with Jules to create documentation and guide future development.

Download the Python repository from resources and references, upload the apps to GitHub, and ask Jules to summarize and explain each application, with readme content and UML alternatives.

Jules delivers readme files with the requested information, markdown files with sequence diagram scripts rendered as images, and data analysis method and reports used, showing how to merge the branch.

Generate code documentation with Jules by creating docstrings for Python files and markdown for notebooks, running in parallel, without changing logic, and publishing branches for merging.

Explore sample documentation generated by Jules, locate docstrings at the top of the application and within functions, and review structured documentation found in Jupyter notebook files.

Explore AI-powered programming with Jules to generate a coffee sales data analysis in a Jupyter notebook, using a sample CSV subset and a comprehensive readme.

Test with the static data set and verify outputs match the analysis result. Rerun with the real dataset and validate the simple chart lacking predictive analysis, highlighting Jules's effectiveness.

Clean the order date in the pizza sales CSV by handling multiple date formats, using a subset of the rows to identify a single format, and implementing the data-cleaning plan.

Explore how Jules, the ai coding agent, handles data parsing errors, switches data sources to csv, and validates code, including parsing the order date with given formats.

Review the section and learn how Jules can explain repositories, generate documentation from code, including docstrings or javadoc, and create readme files or comments, while handling errors with creative workarounds.