AI Agents for Data Analysis
Budget: $250 – $750 USD
I want a set of AI-powered agents, hosted in a GitHub repository, that focus on data analysis while relieving me from repetitive data-entry and management chores. The datasets they will face are semi-structured—think CSV files peppered with free-text notes, JSON logs, or mixed-format exports from third-party tools.
Core objectives
• Parse and normalise those semi-structured sources.
• Automate the data-entry and clean-up pipeline so I never touch a spreadsheet again.
• Run exploratory and descriptive analysis, then surface results through concise reports or dashboards.
I am language-agnostic, but Python with libraries such as Pandas, spaCy, or LangChain would fit well; feel free to propose alternatives if they serve the same purpose. Everything should live in a well-organised GitHub repo, complete with clear commit history and a README that lets me spin the agents up in minutes.
Deliverables
• Fully functional agent code in GitHub
• Setup instructions and dependency file
• Sample dataset plus demonstration notebook or script
• Brief usage guide showing how new semi-structured files flow through the system and end in an analysed state
Once I can pull the repo, run a single command, and watch the agents clean, structure, and analyse a fresh batch of data, the work will be considered complete.
Core objectives
• Parse and normalise those semi-structured sources.
• Automate the data-entry and clean-up pipeline so I never touch a spreadsheet again.
• Run exploratory and descriptive analysis, then surface results through concise reports or dashboards.
I am language-agnostic, but Python with libraries such as Pandas, spaCy, or LangChain would fit well; feel free to propose alternatives if they serve the same purpose. Everything should live in a well-organised GitHub repo, complete with clear commit history and a README that lets me spin the agents up in minutes.
Deliverables
• Fully functional agent code in GitHub
• Setup instructions and dependency file
• Sample dataset plus demonstration notebook or script
• Brief usage guide showing how new semi-structured files flow through the system and end in an analysed state
Once I can pull the repo, run a single command, and watch the agents clean, structure, and analyse a fresh batch of data, the work will be considered complete.