Daniella's Customer Brain

Dark Horse <> OneAI: Tech and Script Optimization

3 min read

Dark Horse <> OneAI: Tech and Script Optimization

Date: 2026-01-28 | Duration: 68.0 min | Account: dark-horse

External attendees: justin@darkhorse.cpa

Summary

The team confirmed the full integration with HubSpot, addressing access issues and enhancing scheduling through a round-robin system among CPAs. Justin Kurn resolved single sign-on conflicts for key users, while workflows were established to streamline booking accuracy and respect calling time zones. A call-back delay rule was implemented to prevent customer overload, and booking confirmations will be made within a specified time frame. AI agents demonstrated foundational workflows with a focus on refining tone and objection handling, while empathy in communications was emphasized. The team is targeting a February 15 launch, with plans for continuous A/B testing and technical improvements to ensure effective CRM data integration and tracking of interactions.

Key Points

🔗 HubSpot Integration: Full integration confirmed, resolving access issues and enabling smooth AI agent scheduling through a round-robin system. 🤖 AI Agent Optimization: Agents to refine tone and objection handling, adding scripted personality responses for enhanced client rapport. 📞 Call Handling Rules: Calls stop if a meeting is booked; agents will track refusal requests and respond empathetically to sensitive issues. 🚀 Launch Planning: Target set for February 15, with detailed scripts and continuous A/B testing to optimize engagement before launch. 🔧 Technical Improvements: New dropdown fields in HubSpot and AI call summaries will ensure better tracking of interactions and lead management.

Overview

  • HubSpot Integration: Full integration confirmed, resolving access issues and enabling smooth AI agent scheduling through a round-robin system.

  • AI Agent Optimization: Agents to refine tone and objection handling, adding scripted personality responses for enhanced client rapport.

  • Call Handling Rules: Calls stop if a meeting is booked; agents will track refusal requests and respond empathetically to sensitive issues.

  • Launch Planning: Target set for February 15, with detailed scripts and continuous A/B testing to optimize engagement before launch.

  • Technical Improvements: New dropdown fields in HubSpot and AI call summaries will ensure better tracking of interactions and lead management.

Action Items

Justin Kurn Provide detailed tone and script edits based on AI call recordings by Friday EOD (01:06:32) Review received call recordings/transcripts shared by Michael and submit feedback on filler word reduction, script flow, and tone (38:41) Share mapping of forms and scheduler links in Slack/email post meeting (09:11)

Michael Gur Provide call recordings and transcripts to Justin for review and script refinement (38:41) Discuss with Daniela about transcript sharing feasibility and ensure delivery (38:41) Set up dedicated HubSpot AI agent field and define dropdown options for AI call outcomes in coordination with Justin (16:18) Continue tuning AI call scripts to reduce filler words, redundant confirmations, and improve empathy and objection handling responses (01:02:20)

Yochai Levi Assist with coordinating next meeting scheduling and follow-up tasks with Daniela (01:06:14) Support implementation of script tonal changes and AI agent personalization with Justin and Michael (40:24) Ensure new script and workflow adjustments align with calling time rules and business hours for launch readiness (27:47)

General / Team Fix website form duplications causing duplicate form submissions in HubSpot (to be resolved in 1-2 months) (23:00) Monitor AI agent learning curve to reduce latency and filler responses over call volumes (01:02:34) Evaluate CEO feedback on call scripts after Justin provides edited call samples (43:46)

Graph View