All evals
G

Eval directory

Evals for Goodcall

Eval coverage for Goodcall, mapped from its public product surface.

About Goodcall

Goodcall is an agentic voice AI platform that answers and automates inbound customer service and sales phone calls with custom AI phone agents. Agents connect to knowledge sources, CRMs, calendars and business tools to handle appointment booking, lead capture and call routing without engineering work. It is sold per agent on tiered monthly plans, with an analytics dashboard for automation rates, call duration and caller behavior, plus a custom enterprise tier.

Industry

agentic voice AI phone agent for customer service and sales

Use the eval library for Goodcall

We'll build out the full library — runnable test cases with inputs, expected behavior, and pass/fail checks — in your Corsac workspace.

Generate your own →

Coverage map

What would you measure for Goodcall?

6 scoring areas · 24 capabilities mapped · grounded in 8 cited pages

Every eval set is graded on

  • Adversarial robustness
  • Workflow quality
  • Safety gates
  • Operator quality

Pass/Fail + LLM judge 1–5 · critical severity flags · negative controls

01

Inbound Call Handling & Conversation Control

How the voice agent answers inbound customer service and sales calls: grounding answers in connected knowledge sources, identifying what the caller wants, and staying inside the business's defined scope during a live conversation.

60,835,286+ voice agent interactions www.goodcall.com

Mapped capabilities

4 capabilities

  • Knowledge-grounded answers

    Responds to hours, services, and policy questions using connected knowledge sources rather than improvising.

  • Intent capture across service and sales

    Distinguishes support requests from new-business inquiries and drives each toward the configured outcome.

  • Out-of-scope and unclear-input handling

    Handles requests outside the configured scope, misheard input, and interruptions without fabricating an answer.

  • Multi-turn context retention

    Carries caller-provided details across turns within a single call without re-asking.

02

Appointment Booking & Rescheduling

The scheduling workflow the agent runs against a connected calendar and CRM, covering slot discovery, booking, changes, and conflict avoidance.

Sync with your CRM and calendar to cut booking time by 5X www.goodcall.com

Mapped capabilities

4 capabilities

  • Availability lookup

    Offers only slots the connected calendar actually shows as open.

  • Booking confirmation and detail capture

    Collects required booking details and confirms the appointment back to the caller.

  • Reschedule and cancellation flows

    Moves or cancels an existing appointment and releases the prior slot.

  • Conflict and double-booking avoidance

    Declines to write an appointment onto an already-occupied slot.

Illustrative example

Input
Caller asks to move their Tuesday 2:00pm appointment to Thursday morning. The connected calendar shows Thursday 9:00am open and Thursday 10:00am already booked.
Expected behavior
The agent offers the 9:00am slot and does not offer 10:00am, then confirms the new time back to the caller and releases the original Tuesday 2:00pm booking.

03

Lead Capture & Delivery

Capturing every inbound caller as a structured record and routing it to the destinations the business configured — SMS, email, Google Sheets, or CRM — without manual data entry.

Capture every inbound call and instantly share it via SMS, email, Google Sheets, or your CRM. www.goodcall.com

Mapped capabilities

4 capabilities

  • Structured caller detail capture

    Captures name, callback number, and inquiry type as discrete fields.

  • Multi-destination delivery

    Emits the captured lead to each configured destination with consistent field values.

  • Repeat and duplicate caller handling

    Recognizes a returning caller rather than creating an unrelated duplicate record.

  • Delivery failure visibility

    Surfaces a failed handoff to a destination instead of silently dropping the lead.

Illustrative example

Input
A new caller leaves a service inquiry with their name, callback number, and job type. The agent's configured lead destinations are SMS and the connected CRM.
Expected behavior
The agent reads the callback number back for confirmation before ending the call, then emits one lead record carrying name, callback number, and job type to both SMS and the CRM.

04

Agent Setup & Logic Flows

The no-code configuration surface used to launch and tune an agent in minutes: connecting knowledge sources and business tools, authoring logic flows, and shaping agent behavior.

Launch your custom phone AI agent in just a few minutes — no engineering team required. www.goodcall.com

Mapped capabilities

4 capabilities

  • Knowledge source and tool connection

    Connects knowledge sources, CRMs, calendars, and business tools into an agent.

  • Logic flow authoring

    Defines branching call logic and respects the flow count allowed by the plan.

  • Behavior and workflow tuning

    Adjusts agent tone, phrasing, and workflow steps through the configuration interface.

  • Directory contact management

    Maintains the directory contacts an agent can transfer or route calls to.

05

Call Routing & Human Escalation

What happens when the agent should not or cannot complete the request itself: transferring to a directory contact, routing by intent, and handing off cleanly.

Mapped capabilities

4 capabilities

  • Intent-based routing

    Routes the caller to the correct directory contact or destination for their stated need.

  • Escalation to a human

    Transfers when the request exceeds the agent's configured capability.

  • Handoff context preservation

    Passes what the caller already said into the transfer rather than restarting.

  • Unavailable-destination fallback

    Falls back to capture or message-taking when the routing target cannot be reached.

06

Analytics, Plans & Account Limits

The reporting dashboard and the commercial envelope around it: automation rate, call duration, and caller behavior reporting, plus per-agent tiered plan entitlements, seats, retention windows, and unique-customer overage.

100 unique customers monthly *$0.50/customer after 100 www.goodcall.com

Mapped capabilities

4 capabilities

  • Automation rate and call metrics

    Reports automation rate, call duration, and caller behavior over a selected period.

  • Call and customer detail drilldown

    Exposes individual call and customer records within the tier's retention window.

  • Plan entitlement enforcement

    Applies per-agent limits on logic flows, team members, and directory contacts by tier.

  • Unique-customer cap and overage

    Tracks monthly unique customers against the plan cap and the per-customer overage beyond it.

Coverage is mapped from Goodcall's public pages (8 crawled). Examples are illustrative, not real test cases. The runnable eval library — graded inputs, expected behavior, and pass/fail checks — is built when you request it above.

Frequently asked questions

What do the Corsac evals for Goodcall test?+

The coverage map is generated from Goodcall's own public product surface (agentic voice AI phone agent for customer service and sales): 6 scoring areas — Inbound Call Handling & Conversation Control, Appointment Booking & Rescheduling, and Lead Capture & Delivery, and more — spanning 24 mapped capabilities, each graded on adversarial robustness, workflow quality, safety gates, and operator quality once the library is built.

How are the Goodcall evals scored?+

Every case generated for Goodcall — across Inbound Call Handling & Conversation Control and Appointment Booking & Rescheduling and the other mapped areas — is graded with pass/fail checks plus an LLM judge scoring 1–5 against its expected behavior, with critical-severity flags and negative controls. Only judge-passed evals are published.

How many test cases does the Goodcall library include?+

The full Goodcall library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, Knowledge-grounded answers and Intent capture across service and sales under Inbound Call Handling & Conversation Control); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

How do I run these evals against Goodcall or my own agent?+

Request the library with your work email above. We'll build out all 6 mapped Goodcall areas and set them up in a Corsac workspace, where you can run every test case against Goodcall or your own agent with your own data.