[mindgraph AI]
MINDGRAPH @ AI
CLASSIFIED

root@mindgraph:~# cat case-studies/field-to-dispatch-operations-platform.md

Field-to-Dispatch Operations Platform

DOCUMENT: Field-to-Dispatch Operations Platform

INDUSTRY: Agriculture

PROJECT: Export Operations ERP

CLASSIFICATION: CASE STUDY // PUBLIC RELEASE

SUMMARY:

A role-based mobile ERP that replaces free-text WhatsApp logging across a banana export operation's entire field-to-dispatch chain.

DETAIL:

14 structured mobile forms with mandatory fields, camera-only capture, GPS/timestamp embedding, automatic recovery/box/volume derivation, 9 roles, and 10 live dashboards — including an automated daily production workbook.

IMPACT:

WORKFLOW STEPS DIGITIZED: 14 forms, 9 roles

16 / AGRICULTURE / EXPORT OPERATIONS ERP

A role-based mobile ERP that replaces free-text WhatsApp logging across a banana export operation's entire field-to-dispatch chain.

Client: Banana export operation, Western India

Tags: Agriculture · Export Operations · Mobile ERP · Role-Based Access · Dashboards

// Problem

Every stage of the harvest-to-dispatch pipeline — field visits, rate negotiations, box counts, dispatch records — was logged as free-text WhatsApp messages, formatted differently depending on who typed it. Nothing was searchable, rates couldn't be audited after the fact, and dispatch counts had to be manually reconciled against paper transport notes.

// Our Solution

An ERP covering the full field-to-dispatch chain, from vendor mapping through cold storage arrival, as one structured system. Fourteen operational steps that used to be free-text messages are now mobile forms with mandatory fields, camera-only capture, and automatic GPS and timestamp embedding. The system computes recovery percentages, box totals, and volume conversions on its own rather than by hand, and surfaces the results across ten live dashboards.

// Technical Details

Structured Field Capture

Every one of the fourteen workflow steps is a mobile form with mandatory fields, not free text. Photo and video capture is camera-only — gallery upload is disabled — and every capture automatically embeds GPS location and a timestamp.

Automatic Derivation & Hidden Business Rules

Recovery percentages, box totals, and volume conversions are computed automatically from structured input. A Village→Location mapping table resolves each field's regional grouping from real historical data, and a location-rules table encodes exceptions confirmed from real records. Anything the system isn't confident about is flagged for a human instead of guessed.

Role-Based Access Control

Nine distinct roles — from field scout through rate negotiator, packing line supervisor, dispatch coordinator, and cold storage QC inspector — each see only the steps relevant to them, with attribute-based access since one employee can carry more than one role.

Operational Dashboards

Ten live dashboards cover everything from an executive daily overview to rate trends, recovery % by vendor and hub, C-Class rejection rates, dispatch-to-arrival mismatches, and a full audit/change log of every record edit.

Built for the Field

Python backend, Node.js/TypeScript application layer, PostgreSQL, Celery and Redis for async processing, fully containerized — designed to work over unstable internet, with English, Marathi, and Hindi language support.