Context

AI commerce impact at Coles

1
NPS

Create better experiences, or turn average ones into less painful ones. A refund request goes from friction to a quick resolution. Complex issues finally get the agent time they deserve.

2
Cost-out

Deflect high-volume post-purchase tickets from the call centre. Human agents get freed up for the complex, high-touch cases that actually need them.

3
Revenue

Revenue generation through better conversion or upsell. Harder to measure, and the proof in market isn't strong yet. Best positioned as a later phase, once the post-purchase foundation is in place.

WISMO is the shortest path to value. The volume is already there. The data is already there — today it reaches customers through a human. Putting it in chat lands on NPS and cost-out at the same time.


Where Is My Order

Four paths through a WISMO chat

Every "where is my order?" conversation lands in one of these four places. The first two self-resolve. The last two hand off to the right team with full context.

1
Lateness, self-resolve
Customer opens with

"Where's my order? It was meant to be here by 2:30, it's 3pm now."

Lorikeet

Looks up live van tracking, apologises, shares the ETA, and issues a goodwill voucher capped by policy.

Outcome

Customer gets an ETA and a $5 apology credit. No human touch required.

2
Delivered, show proof
Customer opens with

"I haven't received my groceries yet — order 1234."

Lorikeet

Confirms delivery, quotes the address and time, and shares the proof-of-delivery photo inline.

Outcome

Customer verifies the drop-off themselves. Human only needed if they dispute the location.

3
Reschedule, driver contact needed
Customer opens with

"I won't be home — can the driver come later?"

Lorikeet

Captures the customer's preferred new window, tags the ticket, and hands off to the delivery team with full context.

Outcome

Delivery team takes over to contact the driver.

4
Suspected misdelivery, escalate
Customer opens with

"That's not my front door. Where are my groceries?"

Lorikeet

Flags suspected misdelivery, sets high priority, posts an internal note, and hands off to the resolution team.

Outcome

Resolution team investigates refund or redelivery. Agent doesn't self-resolve a real dispute.


Compliance

Oversight before and after every message

1
Before the message leaves

Every AI response is checked against your rules before the customer sees it, 24/7. A response that fails a guardrail is rewritten, or the conversation is handed to a human. It never ships.

2
After the conversation ends

Every completed ticket is scored against your rules in Coach. Failures surface by severity, are filterable by policy, and trendable over time.

Lorikeet is purpose-built for complex, regulated industries. ISO 27001:2022 · SOC 2 Type 2 · HIPAA · GDPR. Full documentation, audit reports, and security policies live in the Trust Centre.

Visit the Trust Centre ↗

Stack fit

Works with your stack

Customer (chat, email, SMS)
Salesforce
Lorikeet agent
AI
hands off →
Your team
Human agents
Your APIs
and business logic
Model providers
multiple underlying, no lock-in
Genesys

No Salesforce replacement. No changes to Genesys. Your workflows, your APIs, your business logic — Lorikeet ingests and executes.