Skip to main content

Workflow

Data Analyst provides a conversational approach to data analysis. Users upload data files and ask business questions in natural language, receiving statistical analysis and tables generated through AI-powered Python code execution.

Step-by-Step Workflow​

Step 1: Data Upload​

  • File Upload: Upload CSV or Excel files through the platform interface
  • Data Processing: System automatically analyzes file structure and column types
  • Data Validation: Basic quality checks ensure data is ready for analysis

Step 2: Data Exploration​

Start with basic questions to understand your dataset:

"Show column names"
"How many records are in this dataset?"
"What are the data types for each column?"
"Show me the first 10 rows of data"
"What is the date range in this dataset?"

Step 3: Business Analysis​

Ask specific business questions based on your analysis needs:

Performance Analysis:

"Show the top 15 clients by balance"
"What is the total debit and credit for each client?"
"Who are the top 10 clients by transaction volume?"

Financial Analysis:

"What is the average transaction amount for each client?"
"Which transaction types contribute most to overall volume?"
"Show total transaction amounts by currency"

Step 4: Data Filtering and Segmentation​

Use natural language to filter and focus your analysis:

Time-based Filtering:

"Show all transactions from July 2025"
"What was the highest balance in each month?"
"Give me transaction count for each client on June 5th, 2025"
"Analyze daily transaction volume for May"

Category Filtering:

"Show only EUR currency transactions"
"Filter data for a specific client name"
"Display only Payment transaction types"
"Show transactions above 10,000 in value"

Step 5: Progressive Analysis Building​

Build complex analysis through follow-up questions:

  1. Start Broad: "Show me total debit and credit by client"
  2. Focus Specific: "Show this breakdown for July 2025 only"
  3. Add Dimensions: "Group this by transaction type as well"
  4. Compare Segments: "Compare the top 5 vs bottom 5 clients"

Step 6: Results Download​

  • Table Export: Download analysis results in CSV format
  • Structured Data: Professional formatting with clear headers
  • Ready for Use: Results suitable for further analysis or presentations

Analysis Patterns​

Comparative Analysis Workflow​

"Compare transaction patterns between MUR and USD clients"
"Which clients are most dependent on specific transaction types?"
"How do top 5 clients compare in terms of transaction diversity?"
"Analyze differences between high-volume and low-volume clients"

Time-based Analysis Workflow​

"Show monthly trends in total transaction amounts"
"Which days in May had the highest transaction volumes?"
"Compare Q1 vs Q2 performance by client"
"Calculate month-over-month growth rates"

Best Practices​

Effective Query Construction​

  • Be Specific: Include exact column names and criteria
  • Provide Context: Mention time periods, client names, or specific values
  • Ask Incrementally: Build complex analysis through multiple related questions
  • Use Follow-ups: Build on previous results with additional questions

Progressive Analysis Approach​

  1. Explore: Start with basic data understanding questions
  2. Filter: Focus on relevant time periods, clients, or categories
  3. Analyze: Calculate statistics and identify patterns
  4. Compare: Examine differences between segments or time periods
  5. Visualize: Create charts to communicate findings clearly