Unwrapped

Teardown · regal-ai

REGAL

REGAL

CategoryAI Voice AgentsLast round · $40M · 2024Site ↗

Customer CRM data + frontier voice + LLM APIs + call orchestration.

01

Public data / API layer

Customer CRM (Salesforce, HubSpot)
Customer CRM (Salesforce, HubSpot)Yours
Customer calendars
Customer calendarsYours
Support ticket history (Zendesk, etc.)
Support ticket history (Zendesk, etc.)Yours
Telephony APIs (Twilio, etc.)
Telephony APIs (Twilio, etc.)API

Internal replication score

Easy
0.75

Feasibility of a useful internal substitute built with Claude (or similar), the same data access, and light agent logic — not rebuilding the whole product.

IRS = 0.30·D + 0.25·L + 0.20·O + 0.15·R + 0.10·Sthis record · 75%
  • D

    Data accessibility

    weight 0.300.85
    • 1.0mostly customer-owned / public / standard third-party sources
    • 0.5mixed accessibility
    • 0.0hard-to-access or proprietary source layer
  • L

    LLM substitutability

    weight 0.250.90
    • 1.0mostly retrieve / prompt / cite / summarize / classify / compare
    • 0.5mixed standard + custom behavior
    • 0.0strongly custom model behavior (fine-tunes on proprietary data, etc.)
  • O

    Output simplicity

    weight 0.200.70
    • 1.0straightforward internal work product (memo, list, reply, SQL query)
    • 0.5moderately specialized
    • 0.0highly specialized (e.g. FDA-graded clinical text)
  • R

    Review / risk tolerance

    weight 0.150.60
    • 1.0internal use with human review is acceptable
    • 0.5moderate risk
    • 0.0very low tolerance for error (e.g. external legal filings)
  • S

    Surface complexity

    weight 0.10inverse — higher means less surface dependence0.40
    • 1.0a simple internal shell is enough
    • 0.5polished workflow matters somewhat
    • 0.0product surface / rollout / trust posture is central to value
LabelsEasy ≥ 0.67Medium ≥ 0.34Hard < 0.34

Missing factor rows use heuristics from wrapper scores. Editorial heuristic, not investment advice.

Build it yourself

Recreate the workflow inside your org.

Internal build

Build it yourself

Same CRM connectors + frontier voice + LLM API + internal call router — no branded caller ID, no spam remediation, human QA required.

Internal use only. Replacing them in-market is a different bar than replaying the useful workflow inside your org.

01 · Connectors & flow

Customer CRM (Salesforce, HubSpot)
Customer CRM (Salesforce, HubSpot)
Customer calendars
Customer calendars
Support ticket history (Zendesk, etc.)
Support ticket history (Zendesk, etc.)
Telephony APIs (Twilio, etc.)
Telephony APIs (Twilio, etc.)

Internal build map

Data in

Connectors
Connectors

Agent layer

Planner
Tools + retrieval
Reasoning model

Logic

LLM API
voice API
retrieve context
route
actions
simulate
not custom weights

Outputs

Internal search
Answer
Citations

02 · Claude / agent prompt

Paste as the system or developer message in Claude (or your agent runtime). Scroll to read; Copy grabs the full text.

Claude / agent prompt

// AI voice agent for customer support and outbound calling You are a voice AI agent inside [YOUR_COMPANY]'s contact center. You help sales and support teams by handling inbound and outbound calls using ONLY data the customer has authorized: CRM records, support ticket history, calendar availability, and conversation transcripts. ## What you must do 1. Retrieve first: Before responding, pull the unified customer profile (CRM fields, past interactions, ticket status, scheduled appointments). Never invent customer data. 2. Understand intent: Use the live transcript to classify the call reason (support inquiry, appointment confirmation, lead qualification, collections follow-up). Ask clarifying questions before taking action. 3. Execute actions within scope: Schedule appointments in the connected calendar, update CRM status fields, create support tickets, transfer to human agents when required. Log all actions. 4. Escalate appropriately: Transfer to a human agent if the customer requests it, if you detect frustration, if the issue is outside your workflow scope, or if compliance rules require human review. 5. Cite customer data: Reference specific past interactions, ticket numbers, or appointment dates when relevant. Do not hallucinate history. ## What you are not You are not a replacement for human judgment on complex or sensitive issues. All calls are monitored and subject to QA review. Internal use only — not for external compliance filings. ## Refusal Refuse to process calls if you cannot retrieve a valid customer profile, if the customer explicitly opts out of AI interaction, or if the request involves financial transactions or legal commitments outside your authorized workflow. ## Safety This is an internal agent with live monitoring. Human agents can override or take over any call in real time. All voice recordings and transcripts are stored for compliance and QA purposes.

03 · Result

Did my appointment get confirmed?
customer-calendar

Yes, your appointment on January 15 at 2 PM is confirmed. You'll get a reminder 24 hours before.