This project contains two experiments designed to explore and compare the capabilities of an AI Agent (AutoGPT) with a manually guided data analysis workflow. The Titanic dataset is used as the case study.
Objective: Experience the autonomous task planning and execution capabilities of an AI Agent.
Description: AutoGPT is given the goal:
"Please analyze this dataset and generate a report containing key statistical information and visualizations."
AutoGPT then automatically:
-
Loads the Titanic dataset
-
Calculates descriptive statistics
-
Plots relationships (e.g., survival rate vs gender)
-
Summarizes findings into a report
This experiment highlights the strengths and weaknesses of autonomous data analysis performed by an AI Agent.
Objective: Develop an understanding of the capabilities and limitations of AI Agents by comparing them with a traditional manual analysis workflow.
Description: You manually perform the same analysis tasks and compare the results with AutoGPT:
✔️ Steps taken
✔️ Time consumed
✔️ Depth of insights
✔️ Accuracy and correctness
✔️ Efficiency and clarity
This comparison helps reveal where AI Agents can genuinely improve productivity and where human judgment is still needed.
This project demonstrates:
How AutoGPT behaves as an autonomous data analyst
How its generated report compares to a human-made report
The practical strengths and current challenges of AI Agents in data analysis
Useful for learning about AI Agents, autonomous workflows, and human–AI collaboration in data science.