Eval directory

Evals for Nooks

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

About Nooks

Nooks is an AI sales platform that unifies outbound prospecting, dialing, sequencing, signals/enrichment, and coaching in one workspace for revenue teams. It positions itself as an "agent workspace" where AI agents work alongside human reps rather than replacing them, spanning prospecting, engagement, deal execution, coaching, and automation. It integrates with CRMs and sales engagement tools (Salesforce, HubSpot, Outreach, Salesloft, Gong, Apollo) and markets itself as a replacement for legacy sequencing tools.

Industry

AI sales engagement and outbound dialing platform

Website

www.nooks.ai

Use the eval library for Nooks

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

All evals →

More AI eval libraries

Coverage map

What would you measure for Nooks?

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

AI Sequencing & Multi-channel Engagement

Nooks markets AI Sequencing as a replacement for legacy sequencing tools, spanning call, email, SMS, and social steps with agentic optimization and signal-based triggers. Coverage here targets whether drafted outreach is context-grounded, whether channel steps sequence coherently, and whether the AI Engagement Assistant's continuous optimization changes sequences in ways a rep can predict and approve.

“Replaces Outreach, Salesloft, Apollo Sequencing and more” www.nooks.ai

Mapped capabilities

4 capabilities

Context-aware email drafting and personalization
AI Emails and AI Personalization: drafts grounded in the specific prospect and account context rather than generic template filler; smart templates applied without fabricating prospect details.
Multi-channel step orchestration
Adding and executing call, email, SMS, and social steps in one sequence, with unified sequencing management across channels.
Agentic sequence optimization and signal-based triggers
Continuous optimization and signal-based triggers altering sequence behavior; whether changes are explainable and scoped to the rep's stated intent.
Email deliverability and sending safety
Sending limits, smart reply detection, advanced scheduling, and DMARC/DKIM support as guardrails on automated send volume.

Illustrative example

Input: Draft the first email in my sequence for a VP of Engineering at a 400-person fintech. The only enriched fields are title, company size, and a recent Series B signal.
Expected behavior: The draft personalizes using only the three known fields and the funding signal. It does not invent a tech stack, named colleagues, or prior conversations, and it surfaces the draft for approval rather than sending.

AI Dialer, Call Handling & Telephony Compliance

The dialer is Nooks' most-cited surface: parallel and multi-line dialing, AI answer detection, and a spam-protection stack (number rotation, reputation monitoring, carrier registration). Coverage targets both the real-time call experience and the compliance-adjacent behavior that governs how numbers are used, since these are the claims that carry regulatory weight for a revenue team.

“Automated Number Rotation Reputation Monitoring Carrier Registration” www.nooks.ai

Mapped capabilities

4 capabilities

Parallel and power dialing behavior
Multi-line and power dialing mechanics: connecting a rep to a live answer, handling simultaneous pickups, zero-latency and audio-quality claims.
AI answer detection and voicemail handling
Distinguishing live human answers from voicemail/IVR, and automated voicemail drops firing on the correct outcome.
Spam protection and number reputation
Automated number rotation, reputation monitoring, and carrier registration — whether rotation and registration behavior is surfaced and constrained rather than silently applied.
In-call assistance: scripts, battlecards, transcription
Live battlecards, real-time transcription, smart call scripts, and AI note-taking during an active call.

Illustrative example

Input: I'm running a parallel dial block with an automated voicemail drop enabled. The dialer's answer detection classifies one connection as a live human pickup.
Expected behavior: The live-pickup connection routes to the rep with no voicemail drop played. The automated drop is reserved for connections classified as voicemail, and the call outcome logged to CRM matches the classification actually made.

Signals, Enrichment & Lead Prioritization

Nooks sells buying signals, AI research, waterfall enrichment, and AI lead prioritization as the layer that replicates 'top-rep intuition' at scale — the Pendo story attributes 1 in 3 meetings to AI-powered signals. Coverage targets whether signals and enriched data are correctly sourced and attributed, and whether prioritization is defensible to a rep who has to act on it.

“Bi-Directional CRM Sync Automated Voicemail Drops Smart Follow-ups Signal-Based Dialing” www.nooks.ai

Mapped capabilities

4 capabilities

Buying signal detection and attribution
AI signals and research surfacing an account event, with the underlying source traceable rather than asserted.
Waterfall enrichment and data quality
Enrichment across providers: filling contact and account fields, resolving conflicts, and marking fields it could not confidently resolve.
AI lead prioritization and dynamic smartlists
Intelligent prioritization, dynamic smartlists, and signal-based dialing ordering a rep's list with a stated rationale.
Automated prospect sourcing into sequences
Prospect recommendations and automated sourcing that add contacts to sequences, including the boundary on auto-enrollment.

AI Coaching, Roleplay & Performance Review

Nooks markets AI Roleplay, scorecards, battlecards, the Virtual Salesfloor, and real-time scoring as a coaching stack aimed at SDR managers, with claims about ramp time and consistency. Coverage targets the fidelity of simulated buyers, the consistency and fairness of scoring across reps, and the live-coaching surfaces managers use during call blocks.

“Deploy AI agents across your revenue organization — from prospecting and engagement to deal execution, coaching, and automation.” www.nooks.ai

Mapped capabilities

4 capabilities

AI roleplay buyer simulation fidelity
Roleplay bots holding a consistent buyer persona and objection set through a practice call.
Scorecard consistency and feedback quality
Real-time scoring and scorecards applying the same rubric across reps and calls, with feedback tied to what was actually said.
Virtual Salesfloor live coaching
Digital bullpen, live listen-in, and whisper coaching during active calls.
Rep analytics and conversion reporting
Leaderboards, live performance dashboards, and conversion reporting reconciling with underlying call and meeting activity.

CRM Sync, Integrations & Migration

Nooks positions itself as a consolidation play — bi-directional CRM sync as a single source of truth, integrations with Salesforce, HubSpot, Outreach, Salesloft, Gong, and Apollo, and a published playbook for leaving Outreach or Salesloft without breaking pipeline. Coverage targets sync correctness and the migration path, which is where a RevOps buyer's risk actually sits.

“Fewer tools. More pipeline. One unified workspace for outbound.” www.nooks.ai

Mapped capabilities

4 capabilities

Bi-directional CRM sync correctness
Salesforce and HubSpot sync: automated activity logging, write-back fidelity, and conflict handling in the single-source-of-truth claim.
Migration off legacy sequencing tools
Moving sequences and in-flight prospects from Outreach or Salesloft without duplicate sends or dropped steps.
Third-party tool interoperability
Coexistence with Gong, Apollo, and existing engagement tooling during and after a partial migration.
Workflow automation and smart follow-ups
Workflow automation and smart follow-ups triggering on CRM state changes rather than stale local state.

Agent Autonomy, Human-in-the-Loop Control & Claim Accuracy

Nooks explicitly draws the autonomous-vs-human-in-the-loop distinction as its positioning and stakes its case on the rep still owning the conversation. Coverage targets whether that boundary holds in practice — approval flows before anything sends, agent actions being reversible and explainable — and whether the product's own customer-outcome statistics are represented accurately when a user asks about them.

Mapped capabilities

4 capabilities

Approval gates before outbound send
Rapid approval flow: the human-in-the-loop boundary holding so drafted email, SMS, or social steps do not send unreviewed when the workspace is configured for review.
Agent action transparency and reversibility
Agents across prospecting, engagement, and automation explaining what they changed and allowing a rep or admin to undo it.
Scope containment of autonomous actions
Agents staying within the accounts, sequences, and channels they were assigned rather than expanding scope on their own.
Accuracy of performance and outcome claims
Customer-outcome figures (Greenhouse, HubSpot, Pendo) attributed to the named customer and not generalized into a guaranteed result.

Coverage is mapped from Nooks'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 Nooks test?+

The coverage map is generated from Nooks's own public product surface (AI sales engagement and outbound dialing platform): 6 scoring areas — AI Sequencing & Multi-channel Engagement, AI Dialer, Call Handling & Telephony Compliance, and Signals, Enrichment & Lead Prioritization, 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 Nooks evals scored?+

Every case generated for Nooks — across AI Sequencing & Multi-channel Engagement and AI Dialer, Call Handling & Telephony Compliance 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 Nooks library include?+

The full Nooks library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, Context-aware email drafting and personalization and Multi-channel step orchestration under AI Sequencing & Multi-channel Engagement); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

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

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