01 · VOICE
Warm, confident delivery
Emily maintains an approachable sommelier tone without sounding scripted or overly formal.
ARC Liquor Harvey · Leadership Demonstration
New Emily turns a wine question into a complete customer experience.
A more natural voice, verified Harvey wine choices, less dead air, and a thoughtful second text that feels like a sommelier followed up personally.
Improved voice flow
Verified inventory
Two-stage SMS
Personal recommendation
CRM summary
Listen to the pork tenderloin test call
2 minutes 9 seconds · caller asks for two local wines under $35 · Emily completes the full journey.
2:09
Complete test call from greeting through serving-tip offer.
≈1.8s
Longest detected quiet gap—no extended dead-air moment.
2
Verified local wine choices with price, SKU and pairing logic.
2 texts
Practical choices first; personal sommelier follow-up second.
What leadership should listen for
A conversation that sounds helpful—not automated.
The improvement is not one feature. It is the way voice, verified data, timing and follow-up work together.
02 · PACING
Short conversational gaps
Responses arrive with natural breathing room while avoiding the long airplay pauses found in earlier tests.
03 · INTERACTION
Useful preference question
She hears “sweet and tangy,” compares the verified choices and explains why Riesling is the strongest fit.
04 · FOLLOW-UP
A human-feeling second touch
The first text is practical. The second feels personal—one clear bottle, one reason, tied to the caller’s sauce.
The customer journey
One call. Four connected moments.
1. Understand
Pork tenderloin, local wines and a maximum of $35 per bottle.
2. Verify
Emily returns only bottles supported by the Harvey inventory foundation.
3. Send choices
Names, prices, SKUs, pairing notes and the Harvey pickup path arrive by text.
4. Personalize
After hearing “sweet and tangy,” Emily sends her personal pick: Lunessence Riesling.
Customer proof
The two-text experience changes the feel of the service.
The customer receives something useful immediately, followed by a shorter message that reinforces Emily’s judgment and personality.
Text 1 · Verified choices
The practical message
LA FRENZ Viognier and LUNESSENCE Riesling, both within budget, with enough detail for the customer to compare and shop.
Text 2 · Personal pick
The sommelier moment
Emily chooses the Riesling for the sweet-and-tangy sauce and explains the role of bright acidity and aromatic fruit.
“Consistency creates predictability. Predictability creates scalability.”
The leadership opportunity behind New Emily.
Operational proof
The value continues after the call.
Emily leaves a usable record for the team: what the customer wanted, what she recommended, and the reasoning behind the final choice.
Customer context captured
The AI Call Summary records the dish, budget, exact wines, prices, sauce preference and final recommendation.
Concise workflow summary
A clean team-facing note makes the outcome understandable without replaying the recording.
Detailed call evidence
The automated note preserves timing, call duration and the factual recommendation record for review.
Honest test readout
Strong result, with one final conversational refinement.
What is working now
The call is complete, natural and grounded in verified product data. Emily understands the meal and budget, explains two relevant choices, follows the caller’s sauce preference, sends both messages and closes warmly.
What the next prompt improves
The choices should be sent immediately after consent and first name—before comparison—and the delivery confirmation should be spoken only once.
“Absolutely — let me send those verified picks to you.”
