Let’s Talk

PandasAI: How to Get the Most Out of This Smart Data Assistant at Work

Introduction

In today’s data-driven workplace, professionals often find themselves juggling spreadsheets, reports, and databases—searching for insights that can guide decision-making. While Python’s pandas library has long been a favorite for data manipulation and analysis, it requires a fair amount of coding skill. Enter PandasAI—an innovative extension that integrates artificial intelligence into pandas, allowing users to interact with their data through simple, natural-language commands.

Instead of writing complex lines of code, you can now ask questions like “What were the top five products sold last quarter?” and PandasAI will generate and execute the appropriate pandas operations for you. This combination of AI and analytics makes PandasAI a game-changer for office users, analysts, and anyone who regularly works with data.

Uses of PandasAI

With PandasAI, users can communicate directly with their data. Instead of recalling specific pandas functions, you can ask plain-language questions such as “Show me the average revenue per region” or “Compare sales for Q1 and Q2.” The AI understands and translates your request into Python code, performs the computation, and returns results instantly. This drastically reduces the learning curve for those less experienced in coding.

PandasAI enables faster exploration of datasets. For instance, you can quickly identify trends, outliers, or patterns by simply typing “Which departments had the highest growth rate this month?” This feature is particularly useful for time-sensitive reporting or when making quick business decisions based on emerging data patterns.

Data cleaning can take up to 80% of an analyst’s time. PandasAI can suggest ways to handle missing values, detect anomalies, or even apply corrections automatically. This saves hours of manual work, especially when dealing with inconsistent or incomplete data sources.

Visual storytelling is a critical part of business communication. PandasAI allows you to request visualizations directly—“Plot total revenue by product category” or “Create a histogram of customer ages.” It automatically generates relevant charts, enabling users to create presentation-ready visuals without switching between tools.

Non-technical users often struggle to interpret raw datasets. PandasAI acts as a bridge between teams by allowing marketing, sales, or operations staff to query data independently and extract insights, reducing reliance on data teams and improving workflow efficiency.

PandasAI can also support predictive analytics tasks, helping managers anticipate trends like future sales or expected expenses. When paired with AI models, it becomes a lightweight decision-support system that improves business planning.

Tips on How to Use It

Tools That PandasAI Can Help With

Step-by-Step Guide: How and Where to Use PandasAI

Step 1: Install and Set Up – Install PandasAI using ‘pip install pandasai’. Import pandas and PandasAI, connect your API key, and initialize the AI engine.

“pip install pandasai”

Then in Python, write:

“import pandas as pd

from pandasai import PandasAI

from pandasai.llm.openai import OpenAI

llm = OpenAI(api_token=”YOUR_OPENAI_API_KEY”)

pandas_ai = PandasAI(llm, conversational=False)”

Step 2: Load Your Data – Import your dataset from a CSV or Excel file using pandas. Ensure your columns are properly labeled.

“df = pd.read_csv(“sales_data_q3.csv”)”


Step 3: Ask a Simple Question – Prompt PandasAI with a query like “What is the average revenue by region?” It will return the correct result automatically.

“result = pandas_ai.run(df, prompt=”What is the average revenue by Region?”)

print(result)”

Step 4: Request a Visualization – Ask for a visual summary, for example, “Plot a bar chart of average revenue by region.”

“res_plot = pandas_ai.run(df, prompt=”Plot a bar chart of average revenue by Region, with colors by CustomerSegment.”)”

Step 5: Perform Data Cleaning Tasks – Ask PandasAI to fill missing values or summarize grouped data effortlessly.

“clean_df = pandas_ai.run(df, prompt=”Fill missing values in UnitsSold with the median, then show me total units sold by CustomerSegment.”)”


Step 6: Combine Multiple Datasets – Analyze data across different sources by asking multi-table queries.

Step 7: Apply It in Your Workflow – Use PandasAI for recurring reports, ad-hoc analyses, or data validation before meetings.

Suppose you have sales_df, customer_df, and region_df.

Then Write:

“combined_result = pandas_ai.run([sales_df, customer_df, region_df],

                                 prompt=”Which regions had the highest repeat purchases by premium customers in the last 12 months?”)”

Conclusion

PandasAI redefines how professionals interact with data. By merging natural language input with pandas’ analytical power, it opens new possibilities for faster insights and efficient workflows. Whether you’re a data analyst, manager, or project lead, PandasAI helps automate repetitive tasks and enables informed decision-making. Used responsibly and with human oversight, it can transform how your organization handles data and analytics.

References

Author
Nafiz Imtiaz

Subscribe to Our Newsletter

Subscribe to DCITM’s newsletter for insights on AI, automation, and data innovation. Get industry updates and exclusive opportunities delivered to your inbox.