- Discovery and configuration of the approved scorecard
- Processing of eligible recordings against the configured criteria
- Representative testing and human-review requirements for the selected use
Product detail · Scorecard-specific
Apply the same approved scorecard to eligible calls.
Review eligible calls against consistently defined criteria without treating a generated score as an automatic or authoritative decision.
Product record reviewed August 5, 2026. Source: Andy Lauppe — Owner/Founder; exact scorecard review required.
Structured product record
Best fit
- Teams with a repeatable scorecard and a defined review purpose
- Workflows where each criterion can be tied to observable evidence in an eligible recording
- Customers prepared to test scoring behavior, handle exceptions, and retain accountable human review
Selected scope
Configure before scoring
- Customer-defined criteria and weighting subject to feasibility review
- Approved reporting, delivery, or downstream workflow under separate configuration
- Client-decided provider and privacy options based on requirements
Discovery dimensions · not a default scorecard
Define criteria that a reviewer can understand and test.
These examples are possible dimensions to discuss. They are not included criteria, guaranteed detections, or a statement that subjective qualities can be scored reliably.
Were defined greeting, identification, required-question, required-language, or next-step criteria met?
Did the call contain the customer-defined evidence for an appointment, sale, escalation, or other approved outcome?
Does the recording contain evidence that should be presented to an accountable reviewer under the approved scorecard?
Approval-gated sequence
Keep criteria, processing, review, and action separate.
- 01
Eligible recording
A separately configured source creates a recording under the customer’s approved notice, consent, purpose, and policy.
- 02
Configured scorecard
The approved workflow evaluates eligible recordings against defined criteria, weighting, and evidence rules.
- 03
Test and review
Accountable people examine supporting evidence and representative errors, exceptions, false positives, and false negatives.
- 04
Permitted use
A reviewed score may support only the approved workflow. The score does not itself authorize consequential action.
Client-decided configuration
Provider and privacy choices follow the requirements.
Provider and privacy configuration are client-decided during discovery. The default access model is Public API models, and OpenAI direct is the default provider. Available alternatives include Novita.AI and AWS Bedrock.
Privacy options are available as needed based on client requirements. That statement does not identify or promise a particular privacy architecture or control.
Review AI data-use boundariesRelated, but separate
A scorecard does not include every recording or analysis service.
Standard transcription and summarization are separate. A scoring workflow is built around approved criteria and evidence.
Extraction or classification for a business-specific question is separate from a repeatable scorecard.
Recording movement, destinations, email, webhooks, storage, reports, and integrations require separate configuration and approval.
Before activation
Define evidence, exceptions, and permitted use.
- Eligible recordings from a separately configured source
- Customer-approved recording notice, consent, authorization, purpose, and internal policy
- Defined criteria, evidence, weighting, exceptions, acceptable errors, reviewers, and permitted use
- Defined provider, processing, destination, privacy, security, retention, deletion, access, and delivery requirements
Verified billing structure
Base fee plus processed minutes.
Recurring base fee plus a per processed minute charge; exact rates, minimums, allowances, and other charges require a confirmed scorecard and proposal. This page publishes no price, range, tier, per-seat model, minimum, allowance, included volume, processing-time commitment, or service level.
Consequential-use boundary
Do not turn a score into an unchecked decision.
A scorecard must define reviewer access to evidence, testing, exceptions, escalation, permitted use, and appropriate legal or policy review. No employment, customer, safety, financial, medical, or legal decision should be treated as automatically authorized by a score.
Common questions
Keep every answer tied to the approved scorecard.
Is there a standard Heritage scorecard?
No default scorecard is published. Criteria, evidence, weighting, exceptions, reviewers, and permitted use are defined for the selected workflow.
Are scores always correct?
No. Scores and supporting outputs may be incomplete or wrong and require representative testing and human review.
Can a score make an employment or customer decision?
A score does not itself authorize consequential action. People remain accountable for review, context, policy, legal guidance, and the decision.
Where is data processed and retained?
No universal answer is published. Provider, processing, destination, retention, deletion, training, human-access, security, and contractual details require configuration-specific confirmation.
What does it cost?
The verified structure is a recurring base fee plus a per processed minute charge. Exact rates, minimums, allowances, and other charges require an approved scorecard and proposal.
Related solution