Data Intelligence & Analytics Suite

Information Extraction & Performance Metrics

Data Intelligence & Analytics Suite

Client: Midwest Shooting Center
Module: Information Extraction & Performance Metrics
Date: December 13, 2025

1. Overview

The Solved Group ecosystem separates data into two distinct streams:

Information Extraction (Live): Pulling specific details (Names, Dates, Intents) from the conversation to populate your CRM and Booking tools.

Custom Evaluations (Post-Call): Using AI to "grade" the conversation against your Standard Operating Procedures (SOPs) to generate pass/fail metrics.

2. Information Extractor (Live Data Capture)

The Information Extractor is configured to listen for specific "Variables" during the call. Unlike a simple transcript, this tool formats messy conversation into clean, structured data that can be used by your downstream software.

We categorize these variables into five types:

A. Categorical Select (Drop-Downs)

Function: Forces the AI to categorize the caller into a pre-defined bucket.

MSC Configurations:

  • call_intent: [ Membership_Inquiry, Lane_Booking, Gunsmithing, Class_Signup, Complaint ]
  • shooter_experience: [ First_Timer, Novice, Competitor, Law_Enforcement ]
  • booking_status: [ Confirmed, Rescheduled, Cancelled, No_Show_Recovery ]

B. Boolean (True/False Flags)

Function: Binary flags to trigger immediate logic or CRM tags.

MSC Configurations:

  • risk_solo_renter: True if caller tried to book alone (triggers Safety Flag).
  • has_valid_id: True if caller confirmed possession of Gov ID.
  • is_upsell_opportunity: True if caller agreed to hear about the VIP tier.

C. String (Text Fields)

Function: Captures specific names, notes, or descriptions.

MSC Configurations:

  • firearm_model: Specifics for Gunsmithing (e.g., "Glock 19 Gen 5").
  • referral_source: Where they heard about MSC (e.g., "Radio ad", "Friend").
  • cancellation_reason: The specific "Why" given for a missed appointment.

D. Numerical (Counts & Values)

Function: Mathematical values for capacity planning.

MSC Configurations:

  • party_size: Number of guests (Impacts lane allocation).
  • estimated_arrival: Minutes until arrival (for "Speed" agents).

E. DateTime (Scheduling)

Function: Converts "Next Tuesday around 2" into standard ISO timestamps.

MSC Configurations:

  • appointment_start_time: Synchronized with facility timezone.

3. Custom Evaluations (The "AI QA Team")

While Extraction gets data out, Evaluations assess performance. We set up specific criteria, and the AI reviews the transcript after every call to grade the agent.

This replaces the need for a human manager to listen to call recordings.

Metric 1: Script Compliance Rate

Goal: Ensure agents are following the "Midwest Way" scripts.

Evaluation Prompt: "Did the agent explicitly mention the 72-hour cancellation policy?"

Output: Pass / Fail percentage.

Metric 2: Objection Handling Success

Goal: Measure sales effectiveness.

Evaluation Prompt: "If the user mentioned 'Price', did the agent pivot to the $14.99 Retail Membership value proposition?"

Output: Pass / Fail.

Metric 3: Safety Guardrail Integrity

Goal: Liability protection.

Evaluation Prompt: "Did the agent authorize a rental for a solo shooter?"

Output: Critical Fail (Triggers immediate alert to Ops).

Metric 4: Appointment Conversion

Goal: Track "Gym Model" recurring bookings.

Evaluation Prompt: "Did the call result in a confirmed calendar booking?"

Output: Conversion Rate %.

4. Dashboard Analytics & Reporting

The Solved Group Dashboard aggregates the data above into high-level operational views.

A. Volume & Cost Metrics

Total Calls: Inbound vs. Outbound breakdown per location (Lima vs. Detroit).

Average Duration: Helps identify efficiency issues. (e.g., If "Simple FAQ" calls are taking 10 minutes, the script needs tightening).

Cost Per Call: Operational spend tracking.

B. Call Outcome Analysis

We track how calls end to diagnose network or behavioral issues:

  • Completed: Successful natural conclusion.
  • Voicemail: Agent left a message (Outbound campaigns).
  • User Hung Up: Potential friction point or frustration.
  • Agent Hung Up: Technical error or guardrail trigger.

C. Latency & Performance

P50 / P95 Latency: Measures the delay between the user speaking and the Agent responding.

  • Target: <800ms for "Reception" agents (Matt).
  • Target: <1200ms for "Reasoning" agents (Instructor Secretary).

5. Call Logs & Inspection

For deep-dive auditing, the system retains a granular log of every interaction:

  • Audio Recording: Full playback availability.
  • Transcript: Text-searchable logs.
  • Variable State: View exactly what data (Name, Date, Interest) was extracted at each turn of the conversation.
  • Action Logs: "Proof" of work—shows the exact timestamp the Agent sent the API request to your booking software or fired the SMS webhook.

Solved Group | Configuration Data: MSC-FormCall ("Scarlett")

Client: Midwest Shooting Center Module: Info Extractors & Custom Reports Date: December 13, 2025

📊 Set 1: Information Extractors

Configure these in the "Information Extractor" tab to capture CRM data.

1. booking_status

  • Type: Select (Drop Down)
  • Description: The final result of the call regarding the appointment.
  • Options: Confirmed, Callback_Requested, Not_Interested, Disqualified
  • Query: What is the final outcome of the booking attempt?

2. confirmed_appointment_time

  • Type: Date & Time
  • Description: The specific slot agreed upon for the Sample Session.
  • Query: What date and time did the user agree to meet?

3. primary_objection_detected

  • Type: Select (Drop Down)
  • Description: The main reason the user hesitated or pushed back.
  • Options: None, Price, Time_Schedule, Fear_Intimidation, Spouse_Approval
  • Query: What was the main reason the user hesitated?

4. needs_rental_equipment

  • Type: Boolean (Yes/No)
  • Description: Indicates if the user requires a firearm rental.
  • Query: Does the user need to rent a firearm?

5. bringing_guest

  • Type: Boolean (Yes/No)
  • Description: Indicates if the user is bringing a +1 (Upsell opportunity).
  • Query: Is the user bringing someone else with them?

6. sales_temperature

  • Type: Select (Drop Down)
  • Description: The user's enthusiasm level.
  • Options: Hot, Warm, Cold
  • Query: How interested or enthusiastic was the user?

7. lead_specific_notes

  • Type: String (Text)
  • Description: Specific details about needs, fears, or requests.
  • Query: What specific needs or fears did the user mention?

📈 Set 2: Custom Reports (Post-Call Grading)

Configure these in the "Custom Reports" or "Evaluations" tab to grade performance.

1. Conversion Success

  • Metric: Pass / Fail
  • Description: Did the agent achieve the primary goal?
  • Query: Did the agent successfully book a sample session?

2. Safety Compliance Check

  • Metric: Pass / Fail
  • Description: Did the agent enforce 21+ and ID rules?
  • Query: Did the agent mention age or ID requirements?

3. Personalization Score

  • Metric: Pass / Fail
  • Description: Did the agent use the form data (Name/Goal)?
  • Query: Did the agent reference the user's specific goal?

4. Objection Pivot Success

  • Metric: Pass / Fail
  • Description: Did the agent use "Free" to overcome cost fears?
  • Query: Did the agent mention the session is free/no-cost?

5. Gym Model Education

  • Metric: Pass / Fail
  • Description: Did they explain training is ongoing, not one-off?
  • Query: Did agent explain training is recurring/ongoing?

6. Logistics Verification

  • Metric: Pass / Fail
  • Description: Did we get the rental/guest info needed?
  • Query: Did the agent ask about rentals or guests?

7. Tone Alignment

  • Metric: Pass / Fail
  • Description: Was the vibe correct ("The Midwest Way")?
  • Query: Was the agent welcoming and non-intimidating?

Solved Group | Configuration Data: MSC-ConfirmAppt ("Andrew")

Client: Midwest Shooting Center Module: Info Extractors & Custom Reports (With System Keys) Date: December 13, 2025

Here are the exact configurations for Andrew. I have added the System Key Names (msc_andrew_...) to ensure data organization in your CRM and backend.

📊 Set 1: Information Extractors (Live CRM Data)

Use these Keys when mapping variables in SynthFlow.

1. msc_andrew_confirmation_outcome

  • Type: Select (Drop Down)
  • Description: The final result of the confirmation call.
  • Options: Confirmed, Reschedule_Requested, Cancelled, Left_Voicemail
  • Query: What is the final status of the appointment?

2. msc_andrew_equipment_source

  • Type: Select (Drop Down)
  • Description: Tracks inventory needs for the instructor.
  • Options: Bringing_Own_Firearm, Needs_Rental, Unsure
  • Query: Will the user bring their own gun or rent one?

3. msc_andrew_guest_count

  • Type: Number
  • Description: Total additional people joining (for lane capacity).
  • Query: How many guests is the user bringing?

4. msc_andrew_is_shooting_alone

  • Type: Boolean (Yes/No)
  • Description: Safety Flag. Identifies solo shooters for the rental restriction check.
  • Query: Is the user planning to come to the range alone?

5. msc_andrew_waiver_status

  • Type: Select (Drop Down)
  • Description: Did they agree to sign the waiver ahead of time?
  • Options: Will_Sign_Online, Already_Signed, Will_Sign_In_Store
  • Query: Did the user agree to sign the waiver online?

6. msc_andrew_reschedule_time

  • Type: Date & Time
  • Description: If they couldn't make the original slot, when did they move to?
  • Query: What new date and time did the user select?

7. msc_andrew_upsell_interest

  • Type: Boolean (Yes/No)
  • Description: Did they show interest when Andrew mentioned the Guest Pass?
  • Query: Did the user express interest in the guest pass?

📈 Set 2: Custom Reports (Post-Call Grading)

Use these Keys for your Analytics Dashboard.

1. msc_andrew_report_logistics_check

  • Metric: Pass / Fail
  • Description: Did Andrew ask the mandatory equipment question?
  • Query: Did agent ask if user is bringing a gun or renting?

2. msc_andrew_report_guest_upsell

  • Metric: Pass / Fail
  • Description: Did Andrew attempt to expand the party size?
  • Query: Did agent ask if the user is bringing a friend?

3. msc_andrew_report_solo_guardrail

  • Metric: Pass / Fail
  • Description: If solo & renting, did Andrew enforce the rule?
  • Query: If solo/renting, did agent mention the safety policy?

4. msc_andrew_report_waiver_push

  • Metric: Pass / Fail
  • Description: Did Andrew push the digital waiver to save desk time?
  • Query: Did agent tell the user to sign the waiver online?

5. msc_andrew_report_id_check

  • Metric: Pass / Fail
  • Description: Did Andrew remind them to bring a valid Govt ID?
  • Query: Did agent mention bringing a valid ID?

6. msc_andrew_report_reschedule_pivot

  • Metric: Pass / Fail
  • Description: If cancelling, did Andrew fight to keep the lead?
  • Query: If user cancelled, did agent offer a new time?

7. msc_andrew_report_tone_compliance

  • Metric: Pass / Fail
  • Description: Was he helpful and systematic (not pushy)?
  • Query: Was the agent professional and logistical?