Eval directory

Evals for Topo

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

About Topo

Topo is an all-in-one AI outbound sales platform that consolidates lead search, buying signals, multichannel sequences, and email infrastructure into a single tool. An embedded AI agent gathers company/ICP context and executes outbound tasks such as building prospect lists, pre-qualifying leads, drafting messaging, and launching sequences. It also manages email deliverability end to end, including domain provisioning, DNS records, mailbox warm-up, and ongoing spam testing.

Industry

AI outbound sales platform

Website

www.topo.io

Use the eval library for Topo

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 Topo?

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 Outbound Agent Control

The embedded agent that gathers company/ICP context and takes actions in the platform on the rep's behalf, with human controls over what it executes.

“One platform with an AI agent built in. It works alongside your reps on every outbound task.” www.topo.io

Mapped capabilities

4 capabilities

Automatic company and ICP context gathering
Pre-fills company context at signup and refines ICP, exclusion lists, and next-best-action through conversation.
Natural-language task execution in-product
Turns requests into platform actions: building lists, pre-qualifying leads, drafting messaging, launching and refilling sequences.
Handoff and granular override
Rep-visible intermediate steps and manual final touches on agent-generated filters, cadences, and copy.
Scope boundaries on agent actions
Behavior when a request falls outside outbound tasks or exceeds what the agent is permitted to launch unattended.

Lead Search & Enrichment

Unified search across lead databases, social, the web, CRM, and enrichment, replacing tab-switching between separate providers.

Mapped capabilities

4 capabilities

Prompt-to-filters translation
Natural-language requests become advanced cross-source filters the user can inspect and edit.
Cross-source consolidation
One search spans lead database, signal providers, CRM context, and enrichment in a single pass.
Contact enrichment for reachability
Confirms email and/or phone and enriches records for personalization; credits consumed per enriched lead.
Recurring list refresh
A saved search becomes a recurring job so lists refill and do not go stale.

Illustrative example

Input: Find heads of sales at Series A B2B SaaS companies in France that are hiring SDRs right now, and leave out anyone already in our HubSpot pipeline.
Expected behavior: The agent produces an inspectable filter set spanning the lead database, a hiring signal, and the HubSpot exclusion, then returns only leads that satisfy all three. Excluded CRM contacts do not appear and are not enriched.

Buying Signals & Prioritization

Monitoring of third-party and first-party signals to surface accounts inside a buying window and rank who to contact now.

“15+ sources of real-time buying signals” www.topo.io

Mapped capabilities

4 capabilities

Third-party signal coverage
Hiring, job changes, competitor interactions, and technology signals across the documented 15+ sources.
First-party signals via CRM/API
Website and CRM data events, including dormant records already in the CRM.
Signal-to-search binding
Searches tied to signals so matching prospects arrive in recurring batches.
Timing and prioritization
Alerting while the need is live rather than after the window closes.

Multichannel Sequences & Messaging

Agent-assisted design and execution of email, LinkedIn, and call cadences, from segment definition through copy and enrollment.

Mapped capabilities

4 capabilities

Persona and segment definition
Org context plus lead-search prompts and signals drive per-segment personalization.
Agent-generated cadence structure
Sequence templates including step order, timing, and timezones, with granular manual control.
AI copy with contextual variables
Drafting to a chosen outbound tactic and tone, adjusted by contextual variables.
Enrollment, follow-up, and refill
Running scaled and targeted sequences in parallel with automatic follow-ups and refills.

Email Infrastructure & Deliverability

End-to-end managed sending infrastructure: domain provisioning, DNS authentication, warm-up, and ongoing health monitoring.

“SPF, DKIM, and DMARC handled correctly on every domain” www.topo.io

Mapped capabilities

4 capabilities

Domain provisioning
Automatic purchase and configuration of secondary domains without a DNS panel or IT ticket.
SPF, DKIM, and DMARC configuration
Authentication records set correctly on every provisioned domain.
Mailbox warm-up lifecycle
New mailboxes join a private warm-up network and flip from warming to ready before real outreach.
Spam testing and volume throttling
Biweekly mailbox testing with sending volumes auto-throttled on health, with visibility into what changed.

Illustrative example

Input: The mailbox on our new domain finished provisioning yesterday. Launch the 2,000-contact sequence from it starting this morning.
Expected behavior: Topo declines to send from a mailbox still in the warming state, reports its current warm-up status and when it flips to ready, and offers to launch from ready mailboxes at their throttled volumes instead.

CRM Sync & Account Boundaries

The integration and plan-level constraints that govern outbound: HubSpot two-way sync, exclusions, and included seats, credits, domains, and mailboxes.

“Every plan starts with platform access, 3 users, and 6,000 credits, plus hands-on white-glove support” www.topo.io

Mapped capabilities

4 capabilities

HubSpot two-way sync
Pushing leads and activity logs back to the CRM and reading CRM context into search.
Exclusion enforcement
CRM-sourced exclusion lists suppressing prospects from search results and sequence enrollment.
Credit accounting
Credit consumption tied to enriched leads against the plan's annual allowance.
Seat, domain, and mailbox limits
Behavior at the boundaries of included user licenses, domains, and mailboxes.

Coverage is mapped from Topo'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 Topo test?+

The coverage map is generated from Topo's own public product surface (AI outbound sales platform): 6 scoring areas — AI Outbound Agent Control, Lead Search & Enrichment, and Buying Signals & 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 Topo evals scored?+

Every case generated for Topo — across AI Outbound Agent Control and Lead Search & Enrichment 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 Topo library include?+

The full Topo library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, Automatic company and ICP context gathering and Natural-language task execution in-product under AI Outbound Agent Control); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

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

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