Ai Project Template

AI Project Template

[](https://github.com/Sagargupta16/ai-project-template/actions/workflows/ci.yml) [](https://github.com/Sagargupta16/ai-project-template/actions/workflows/codeql.yml) [](LICENSE) [](https://www.python.org/downloads/)

A unified Python starter for **every AI use case** -- classical ML, deep learning, LLM apps, RAG, and agents. One layout, optional extras, opinionated tooling.

Why this template

Most AI templates pick a niche. Cookiecutter Data Science assumes sklearn + notebooks. Production RAG templates assume a FastAPI service with a vector store. This template assumes **your project will span several of those** (ML model → RAG retriever → agent tool → API) and gives each concern its own home without forcing you to install frameworks you don't need.

Quick start

# Install uv (https://docs.astral.sh/uv/)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Install base + the extras you need
make setup                                    # base only
uv sync --extra ml --extra llm --extra rag    # targeted
make dev                                      # everything

# Run
make test              # tests
make api               # FastAPI dev server on :8000
make eval              # offline evals on the golden dataset
make notebook          # Jupyter

Extras

Base install is deliberately lean -- just `pydantic`, `pydantic-settings`, `python-dotenv`, `tqdm`. Everything heavy is an opt-in extra:

Extra	Adds	Use when
ml	numpy, pandas, sklearn, matplotlib	Classical ML, data science
dl	torch	Deep learning
llm	anthropic, openai, tiktoken	Calling LLMs