Thoughtly Office Hours - February 17, 2025

​

Calendarly and GHL Integration [00:02:00]

• Discussion on the new persistent calendar invite link to simplify regular attendance without rebooking.
• Addressed the reactivation of the Calendarly integration and suggestions to work with question prompts.
• Noted the progress towards integrating Google High Level (GHL) with Thoughtly version 1.5, focusing on creating nodes to manage scheduling and its associated actions.

​

Handling Delays on Call Connections [00:06:00]

• Desmond and Tony raised concerns about call agents starting to talk immediately upon connection, often before the recipient has their phone up to their ear.
• Torrey shared the issue’s connection to outbound call interface detection and the latency of language models.
• Proposed adding a delay feature to facilitate proper call pick-up and interaction, with potential plans to improve answering machine detection using alternative technologies like Telnix.

​

Version 1.5 Improvements and Challenges [00:10:00]

• Tony K highlighted challenges with the Thoughtly 1.5 builder regarding conversation flow rigidity compared to the fluidity of 1.0.
• The 1.5 builder is perceived as too linear, leading to less dynamic conversations similar to early IVR systems.
• Torrey explained the purpose of V1.5 as a solution for compliance-driven enterprise clients needing predictable call scripts.
• Announced upcoming V2 builder, combining fluidity and strict actions, resolved many of V1.5’s limitations.

​

Industry Use Cases: Car Dealerships [00:18:00]

• Tony Q inquired about experiences using Thoughtly with car dealerships, expressing interest in transforming BDC operations.
• Explained the common agency structure and the pricing model for agency fulfillment while participating in Thoughtly’s partner network to amplify business opportunities.
• Direction given for connecting with existing partners like Will to assist with implementation.

​

Creating Contacts and Texting Workflows [00:24:00]

• Shane asked about generating contacts during a call and the differentiation between call and text workflows in the new V1.5.
• Advised against creating contacts during calls due to processing time, suggesting post-call automation for tasks not immediately essential.
• Concerning text workflows, suggested building separate agents better suited for text-based interactions to avoid confusion in mixed media interactions with the same agent.

​

Implementing Memory in Conversations [00:44:00]

• Steven explored maintaining conversation context across multiple calls, asking about variable use for conversation continuity.
• Tony K and Torrey proposed utilizing API and metadata to store and reference conversation history.
• Emphasized customizing memory prompts in workflows to better retain essential information and enhance AI responses during follow-up interactions.

​

User Feedback on Version 1.5 Outcomes [00:53:00]

• Savvy Matthew provided feedback on the V1.5’s handling of outcomes, preferring previous simplicity for quick phrase matching.
• Carlos assured that usability issues are acknowledged and solutions are forthcoming, as addressed in the upcoming V2 design.
• Hover-based solutions were discussed to improve outcome displays, with input requested from the broader community.

​

Promoting and Enhancing the Partner Program [00:56:00]

• Matthew discussed implementing Thoughtly with 3CX for small businesses and the potential for enriching the Partner Program.
• Torrey highlighted the Partner Program’s value in providing agencies with fulfillment contracts, emphasizing collaboration within the community.
• Encouraged experienced users to join and support the network, offering support through Thoughtly’s Slack group.

​

Caller ID and Number Management [01:05:00]

• Tony K queried methods to enable single number utilization across multiple agents, focusing on outbound call ID functionality.
• Discussion emphasized potential verification methods and system limitations on assigning numbers across several agents.
• Explored practical solutions like temporary agent creation, verification, and thematic feedback on improvement.