Vantage <> OneAI: Weekly Sync

Date: 2025-09-10 | Duration: 41.11000061035156 min | Account: vantage-point

External attendees: brittanyr@vantagepointsoftware.com, daniels@vantagepointsoftware.com, romanb@vantagepointsoftware.com

Summary

The meeting addressed the current status and preparations for the OneAI system’s launch, confirming agent readiness and awaiting Vantage team approval for scripts and final testing. Voice testing was clarified, with the system correctly assigning voices for A/B testing to maintain validity. Amit demonstrated voice control features, allowing manual selection of specific voices for testing, with a focus on testing the male voice, Brian, due to the predominance of males in the trading industry. The call sequence structure was defined, outlining a total of 15 attempts with specific timing to enhance outreach effectiveness. Quality control measures, including emergency stop buttons and contact prioritization, were discussed, alongside concurrent call management capabilities. The team aims for a live launch within a week, pending successful voice testing and integration setups.

Key Points

📊 Project Status: Agent readiness confirmed; awaiting Vantage team approval for final testing. 🔧 Call Sequence Defined: A structured call sequence of 15 attempts ensures effective outreach. ✅ Quality Control Measures: Multiple control layers in place to ensure effective transfer and manage calls. 📞 Concurrent Call Management: Scalable capacity settings allow for up to 200 concurrent calls. 🚀 Next Steps: Voice testing and integration setups needed for a target launch within a week.

Overview

  • OneAI system is technically ready for launch, awaiting Vantage team approval for script and final testing.
  • Voice testing will prioritize Brian’s male voice due to the 75% male-dominant trading industry, as emphasized by Daniel.
  • Call sequence defined: 15 total attempts over varying intervals to optimize outreach effectiveness and avoid same-hour calling.
  • System handles callback requests automatically, logging them and maintaining contact in sequence until resolution.
  • Salesforce checkbox controls activation of call sequences, ensuring sequences can resume correctly if paused and re-enabled.
  • Quality safeguards in place include an emergency stop button and individual call pausing options for added control during calls.
  • Fresh contact prioritization established at 75% for new leads, focusing on improving outreach to newer prospects.
  • Current call capacity set at 2 concurrent calls, with scalability up to 200 based on agent availability and demand.
  • Predictive throttling will manage call volume automatically, adjusting to agent availability and transfer ratios after baseline data is established.
  • Timeline target set for live launch within one week, pending successful completion of voice testing and transfer queue setup.

Action Items

Daniel Santiago Test different AI voices (Raquel, Aria, Brian, Matilda) to evaluate performance across male/female options (34:24) Set up Ring Central transfer queue and assign agents for call handling (35:00) Conduct internal testing of transfer queue functionality with team members (35:00) Coordinate with Brittany and Roman on final go-live approval process (35:15)

Brittany Review and approve final script for AI agent before go-live (02:02) Coordinate with Roman on checkbox management strategy and internal processes (27:06) Finalize internal discussions on sequence interruption protocols (28:13)

Roman Verify webhook data fields include all necessary callback and sequence information (11:35) Review integration approach with Brittany for checkbox management (27:06) Test concurrent call batch processing capabilities (35:29)

Amit Ben Keep voice experiment disabled and reset to original A/B testing configuration after voice testing complete (08:07) Monitor fresh contact percentage setting at 75% as requested (38:02) Provide support during testing phase and be ready for go-live launch (34:07)

Michael Gur Verify callback requested field is properly included in webhook data (13:04) Ensure sequence and attempt number fields are correctly transmitted (32:47)