📊 Data Science DC – Code Examples

Code exammples from Data Science DC in May 2025.

---

Baseline API (/api)

Run this API to provide data to the lessons

• ✅ FastAPI framework written in Python
• 🔁 SQLite3 database for sample data

🛠 Skill 1 – Create a Python SDK (/sdk)

Develop a Python SDK to interface with your FastAPI-based API. This example demonstrates:

• ✅ Pydantic models for robust data validation and serialization
• 🔁 Exponential backoff with retry mechanisms for fault tolerance
• 🪵 Integrated logging for observability and diagnostics

Ideal for those building internal or public-facing API client libraries.

📂 Folder: /sdk

---

🤖 Skill 2 – Create an AI Agent (/ai_agent)

Construct an interactive AI agent using state-of-the-art tools:

• 🧠 LangGraph for defining structured agent workflows
• ✨ Anthropic APIs (e.g., Claude) for LLM-driven interaction
• 📓 Jupyter Notebooks for iterative development and demonstration

(Also uses SDK from step 1)

A must-see if you're exploring generative AI, reasoning agents, or tool-using LLMs.

📂 Folder: /ai_agent

📦 Skill 3 – Deploy a Machine Learning Model (/ml_model)

Serve a trained scikit-learn model via a production-grade API using:

• ⚙️ ONNX Runtime for optimized inference across environments
• 🌐 FastAPI to expose the model as a performant RESTful service
• 🧪 Example pre- and post-processing code for online inference

This project is a blueprint for transitioning from local notebooks to scalable deployment.

📂 Folder: /ml_model

---