The 3-Day Pilot Rollout Playbook

This operational runbook provides the field methodology for deploying KHOLO into a live South African wholesale distributor without disrupting existing operations or risking order errors.

---

1. The Core Principle: Zero Operational RiskCopy Anchor Link

Wholesale distributors cannot afford downtime, misplaced orders, or incorrect shipments. Therefore, Kholo is rolled out in three distinct phases:

---

2. Day 1: Shadow Mode (Zero Operational Risk)Copy Anchor Link

Objective:Copy Anchor Link

Measure baseline accuracy and capture the distributor's unique customer jargon without touching their live Sage Evolution ledger.

Workflow:Copy Anchor Link

1. Normal Business As Usual: Order clerks continue receiving WhatsApp orders on their current phones and manually typing them into Sage Evolution as they have always done.
2. Kholo In Parallel: All incoming messages are mirrored or forwarded to the Kholo WhatsApp pilot number.
3. End-of-Day Alignment Audit:
   • Export the day's manually created Sage orders.
   • Compare them against Kholo's auto-extracted orders in /dashboard.
   • Measure Baseline Accuracy: Typically $72% - 78%$ on unprimed catalogs.
   • Catalog Gap Analysis: Review the unmatched items. For example, if 15 customers asked for "12.5 whites", add the alias to White Star Maize Meal in Kholo.
   • Result: Before clerks even touch Kholo, the brain is pre-conditioned with their customers' actual dialect.

---

3. Day 2: Hybrid Mode (Clerks Become Supervisors)Copy Anchor Link

Objective:Copy Anchor Link

Transition order clerks from manual typists to reviewers using the Kholo Control Centre.

Workflow:Copy Anchor Link

1. Clerk Stations: Order clerks sit in front of the Kholo Dashboard (http://localhost:XXXX/dashboard).
2. Triaging Orders:
   • Green Orders (95%+ match): Clerks do a 3-second visual check and click "Approve".
   • Amber Orders: Clerks click "Resolve & Teach", select the right variant, and save the customer alias.
3. Batch ERP Export:
   • Instead of direct live API sync on Day 2, approved orders are exported via "Export Batch CSV" every 2 hours and imported into Sage in bulk.
4. Key Metric Tracked:
   • Average order processing time drops from 4 minutes (manual typing) $\rightarrow$ 35 seconds (dashboard review).

---

4. Day 3: Full Live Processing (Automated ERP Creation)Copy Anchor Link

Objective:Copy Anchor Link

Full automated pipeline: WhatsApp message $\rightarrow$ Kholo $\rightarrow$ Sage Evolution $\rightarrow$ Warehouse Pick Slip $\rightarrow$ Customer WhatsApp Confirmation.

Workflow:Copy Anchor Link

1. Live Direct Line: The distributor's official WhatsApp ordering number points directly to Kholo's Meta Cloud API webhook.
2. Automated Flow:
   • High-confidence orders from credit-approved customers are approved with 1 click.
   • Kholo posts the order to Sage Evolution via the Freedom Service REST API.
   • Sage generates and returns the official sales order number (e.g., SO-2026-8454).
3. Warehouse Dispatch:
   • The warehouse dispatch desk opens /dashboard, clicks "View Pick Slip", and prints the slip organized by aisle and bay.
4. Customer Confirmation:
   • Kholo automatically sends a WhatsApp reply to the buyer:

     "Thanks Clerk! Order received (Sage SO #SO-2026-8454). Total: R1,937.85. Packing now for Delivery Route 04."

---

5. Pilot Evaluation KPIs & Sign-Off CriteriaCopy Anchor Link

At the end of Day 3, review these 4 metrics with the distributor's Managing Director: