Datachat 2.0 is an intelligent data analysis agent designed specifically for the telecommunications industry, aiming to help users rapidly extract business insights from raw data, identify potential issues, generate visual reports, and provide actionable recommendations. Users can upload their own data and specify analytical requirements; alternatively, without uploading data, users can ask general knowledge questions (non-Agent mode).
✅ How to Use?
1. Upload Data (Recommended!)
👉 Supports Excel / CSV formats
👉 Maximum of 10 files (Excel files may contain multiple sheets); each file or sheet must not exceed 500,000 rows and 200 columns
🚨 Please unmerge any merged cells. Clearly label column names in the first row—for example, “User ID,” “Plan Type.” For columns such as “Downlink Throughput” or “Distance,” please explicitly indicate units in the column names, e.g., “dl_th_mbps,” “distance_km.” Avoid generic names like “Column1, Column2”!
2. Choose a Mode
🟢 Lite: Fast! Ideal for getting a quick overview
🟡 Plus: Stable! Suitable for in-depth relationship exploration
🔵 Pro: Deep! Longer thinking time, more comprehensive analysis
3. Ask a Question
👉 With data uploaded:
“Analyze why users choose a particular operator’s mobile money service. Note: the file is in French—please translate results into Chinese.”
“What is the relationship between user tenure and ARPU? How do spending behaviors differ between new and existing users?”
👉 Without data uploaded:
“How should 5G base stations be located?” → The system will provide a brief answer using its built-in knowledge (without launching deep analysis).
🚨 Pitfall Avoidance Tips:
× Do not upload tables containing merged cells
× Avoid using generic column names like “Field1, Field2”
