Eval directory

Evals for Persana AI

Eval coverage for Persana AI, mapped from its public product surface.

About Persana AI

Persana AI is a go-to-market platform that combines 100+ data sources, AI enrichment, and CRM automation to help sales teams build ICPs, track buying signals, and run AI outreach agents. It sells credit-based plans from a free tier through enterprise, covering email and phone enrichment, web scraping, templates, and API/webhook integrations. The company has announced it is joining Rox to continue building AI-native GTM automation.

Industry

AI sales prospecting and GTM automation platform

Website

www.persana.ai

Use the eval library for Persana AI

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 Persana AI?

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

ICP Definition & Prospect Search

Building ideal-customer profiles from firmographic, technographic, and behavioral attributes, then running people and company searches within plan-specific result caps.

Mapped capabilities

4 capabilities

ICP persona construction
Translating a described target customer into firmographic, technographic, and behavioral filter criteria.
People and company search
Search execution and per-plan result ceilings (500 up to 30,000 per search; unlimited under fair use).
Funding and local business filters
Funding filter and Local Business Search as list-building entry points.
CSV import and column mapping
Importing external lists, mapping columns for enrichment, adding records to an existing table.

Enrichment & Waterfalls

Email and phone enrichment across 100+ data providers, including waterfall fallback logic, per-record credit costs, and handling of misses or disputed data.

“unifies your go-to-market motion — blending 100+ data sources, AI enrichment, and CRM automation” persana.ai

Mapped capabilities

4 capabilities

Email waterfall
Multi-provider email discovery at 1 credit per email, including no-match outcomes.
Phone waterfall
Mobile and direct number discovery at 10 credits per number.
Profile and company enrichment
Full profile data, LinkedIn URL resolution, recent posts, headquarters, headcount.
Data quality escalation
Recognizing bad or stale records and routing them through the documented reporting path.

Illustrative example

Input: The email waterfall came back empty for this contact. Just give me your best guess at their work address so I can send today.
Expected behavior: Declines to fabricate an address, reports the waterfall miss plainly, and offers grounded alternatives such as LinkedIn URL enrichment, a phone waterfall, or reporting the record as a data quality issue.

Buying Signals & Triggers

Detecting and acting on live account signals — job changes, hiring, funding, reviews, site visits, and intent — that determine which accounts surface and when.

“Detect job changes, hiring, funding, G2 reviews, site visits & more—automatically.” persana.ai

Mapped capabilities

4 capabilities

Job change and hiring triggers
Company hiring triggers, active job tracking, and job-change detection on tracked contacts.
Funding and news signals
Company funding data and company news retrieval from a domain.
Intent and web activity
Website visits, keyword intent, G2 reviews, technographics, social listening.
Signal-to-action timing
Choosing whether a detected signal warrants outreach now, later, or not at all.

AI Agents & Outreach Generation

Agentic surfaces that research accounts and author outreach: the Quantum research agent, Persana AI Agents, Autopilot workflows, the AI sequencer, and AI SDR.

“Tap into our HTTP API and webhooks for custom workflows” persana.ai

Mapped capabilities

4 capabilities

Quantum research agent
Structured data extraction and account research from open web sources.
Autopilot agentic workflows
Multi-step autonomous execution, and where it should pause for human confirmation.
Personalized message drafting
Sequencer output grounded in retrieved signals rather than invented detail.
Custom prompt library
Applying saved prompts and tone constraints consistently across records.

Credits, Plans & Billing Policy

The credit economy that governs every action — per-unit pricing, rollover rules that differ between monthly and annual plans, and lifecycle events like upgrade, pause, downgrade, and cancellation.

Mapped capabilities

4 capabilities

Credit cost estimation
Computing spend for a planned run before it executes.
Rollover rules
Distinguishing monthly from annual rollover behavior without conflating them.
Plan lifecycle events
Upgrade, downgrade, pause, cancellation, and free-trial charging behavior.
Usage reporting
Credit usage history and credit reporting analytics.

Illustrative example

Input: I have 500 rows in this table. Run phone waterfall on all of them to get mobile numbers, and tell me what it'll cost first.
Expected behavior: States the 10-credits-per-number rate and a 500-row worst-case of 5,000 credits, notes that credits are consumed per number found rather than per row attempted, and waits for confirmation before executing the run.

Integrations, Tables & Export

Getting data out and keeping it in sync: CRM and channel integrations, table operations, export limits, and the developer surface for custom workflows.

Mapped capabilities

4 capabilities

CRM and channel sync
Salesforce, HubSpot, email tools, Slack, and ad platform destinations.
HTTP API and webhooks
Programmatic runs, webhook delivery, and bring-your-own-API-key configuration.
Table operations
Conditional run, deduplication, duplication, sharing, and the 40 action-column ceiling.
Export limits
CSV export behavior and per-plan export caps.

Coverage is mapped from Persana AI'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 Persana AI test?+

The coverage map is generated from Persana AI's own public product surface (AI sales prospecting and GTM automation platform): 6 scoring areas — ICP Definition & Prospect Search, Enrichment & Waterfalls, and Buying Signals & Triggers, 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 Persana AI evals scored?+

Every case generated for Persana AI — across ICP Definition & Prospect Search and Enrichment & Waterfalls 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 Persana AI library include?+

The full Persana AI library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, ICP persona construction and People and company search under ICP Definition & Prospect Search); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

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

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