LO-065 [Advanced]: Lead Import CSV Column Automapping Engine

[Kuldeeep18]

Description

When users import a CSV, they must match their spreadsheet headers manually to LeadOrbit columns (First Name, Email, etc.). If the headers don't match exactly, the import fails. We want to build an intelligent auto-mapping engine.

User & Contributor Value

• Contributors: Implementing fuzzy string matching algorithms (e.g. Levenshtein distance) in Python and designing map forms.
• Users: High usability. Saves time by automatically guessing mapped fields for various spreadsheet formats.

Code Locations

• backend/leads/tasks.py
• frontend/leads.html

Implementation Guide

1. Fuzzy Matcher: Write a utility matching input headers against target columns (e.g. mapping "E-mail", "Mail Address" -> "email").
2. Preview Mapping Form: When uploading a CSV, parse the first row on the backend, predict the mapping, and return it to the UI. Display a column mapping confirmation table showing mapped fields.

[manikanta-tamminana]

Hi @Kuldeep18,

I would like to work on this issue under GSSoC.

My implementation plan is:

• Develop a fuzzy header-matching utility to intelligently map CSV column names to LeadOrbit fields using similarity-based matching techniques.
• Support common variations such as "E-mail", "Email Address", "Mail Address", "First_Name", etc., and map them to the appropriate system fields.
• Parse the CSV headers during the upload process and generate predicted field mappings on the backend.
• Return the suggested mappings to the frontend and display them in a user-friendly mapping confirmation table.
• Allow users to review and adjust the suggested mappings before completing the import.
• Ensure the solution remains extensible for future lead fields and custom column names.

This feature will significantly improve the CSV import experience by reducing manual effort and handling a wide variety of spreadsheet formats more intelligently. Could you please assign this issue to me?

Thank you! 🚀

[NikithaDineshKumar]

Hi @Kuldeeep18, I'd love to work on this intelligent auto-mapping engine! Here is my proposed approach for implementing this feature:

1. Backend Updates (backend/leads/tasks.py)
   Fuzzy Matcher Utility: I will implement a string distance matching utility (such as Levenshtein distance or Python's built-in difflib.get_close_matches) to evaluate incoming CSV headers against standard expected target columns (e.g., mapping variations like "E-mail", "Mail Address", or "Email ID" directly to "email").

Mapping Predictor API: I will update or create an endpoint that parses just the first row (the header) of the uploaded CSV, runs it through the fuzzy matcher, and returns a predicted JSON mapping dictionary to the frontend.

2. Frontend Updates (frontend/leads.html)
   Preview Mapping Form: I will design a clean, responsive confirmation table interface that displays the incoming CSV headers next to their automatically guessed LeadOrbit target fields.

Fallback Selectors: I will include dropdown selectors for each row, allowing users to manually override and correct the mapping if a fuzzy match prediction isn't completely accurate before final submission.

Could you please assign this issue to me? Thanks!

/assign