Platform

Describe who you are looking for, import a list or start from a saved audience. The agent turns your request into criteria and builds the dataset.

• Natural language
• Import a list
• Criteria

Reusable audience definitions, prioritized by focus. "We sell to construction companies" becomes "property and land developers in São Paulo, 5 to 200 employees, mid- to high-end", ready to start the next session.

• Focus
• Nice-to-have
• Low priority

The history of your work with the agent: every conversation, the criteria used and each dataset edition, including what changed.

• History
• Dataset editions

Qualified accounts reach your CRM with context, signals and a contact path for pre-sales, sales and customer success. The agent handles the manual work and provides guidance; your team approves and acts.

• Pre-sales
• Sales
• Customer success
• Human in the loop

A supervisor delegates each step to specialized agents, with guardrails for scope, moderation and cost.

• icp-architect
• company-finder
• contact-finder
• enricher
• qualifier
• crm-connector

Public records, marketplaces, directories, websites, maps, purchased databases and the client's CRM.

• Receita Federal e CNPJ
• CVM
• Google Maps
• Marketplaces
• Sites

Clean, normalize, deduplicate, enrich, index and validate until the agreed fill rate is reached.

• Fill rate
• Deduplication
• Enrichment

Agents read each account's website, social profiles, job openings and public records at scale, applying your qualification rules.

Companies discovered and evaluated against your ICP: fit against each criterion, enrichment, contacts and what has been sent to the CRM. Every filter or adjustment is recorded as a dataset edition.

• Fit
• Criteria
• Enrichment
• CRM delivery

From mapped to qualified companies, from extracted to enriched contacts, across all sessions. Every stage shows how much was discarded and why.

• Mapped
• Qualified
• Contacts
• Enriched

Alerts derived from the Workspace's own data, ranked by severity and each with a next step: contacts without verified email or phone, qualified accounts that have never been enriched.

• Critical
• Attention
• Opportunity

Prospect-to-opportunity conversion by channel, throughout the funnel and session by session.

• Conversion by channel
• Per session

Spend per execution and acquisition cost by channel, so you know where your money goes furthest.

• Cost per execution
• CAC by channel

Companies, contacts and deals synced with your CRM. HubSpot today; Salesforce and Pipedrive coming soon.

• HubSpot
• Salesforce (coming soon)
• Pipedrive (coming soon)

Data providers centralized in a single workflow instead of a fragmented stack paid for separately. The agent chooses which one to query for each account.

• Apollo
• Lusha
• Clay
• Econodata
• BigDataCorp
• BrightData
• Firecrawl

Every account reaches your CRM with its qualifying signal, ICP fit and contact path, ready for your team to act.

• Signal
• Fit
• Contact path

The Ontology is the Workspace's schema: the allowed types, what each means and how they connect. Records live in the Graph, each within its type. The account you will approach is a Company record.

• Entities
• Relationships
• Graph

Each type declares its identifiers, such as a company's website domain. This is how the agent knows two records refer to the same entity and avoids duplicate accounts.

• Identifiers
• Domain
• Email
• URL

Typed properties with rules: required fields, allowed options and value ranges. Signals and pain points are types of their own, linked to the account through relationships.

• Properties
• Validation
• Relationships

Changes begin in conversation. The agent proposes new or revised definitions, and any Workspace member can approve or reject them. The Ontology screen is read-only: nothing changes without approval.

• Proposal in the conversation
• Approve or reject

Approval publishes the definitions together as the next Ontology version. Records retain the definition version they were created under; no records are migrated.

• Ontology version
• Definition version

Each Workspace has its own Ontology, reflecting how you sell: campaigns, opportunities, competitors and customer pain points. It grows with each cycle and stays with you.

Clean, transform and cross-reference databases beyond the reach of off-the-shelf tools, in the format your operation uses.

• Transformation
• Cross-referencing databases

Code that retrieves data from sources unique to your market: portals, websites and APIs beyond standard databases.

• Scrapers
• APIs

Custom scores, metrics and models calculated from your Workspace data.

• Scores
• Metrics
• Models

All code the agent writes runs in an isolated virtual machine.

The cycle on one page: funnel, conversion by channel, cost and generated pipeline, ready to explore and filter instead of reading a PDF.

• Funnel
• Conversion
• Pipeline

A before-and-after view of your market and operation: where the accounts are, what is known about each and what has changed.

• Before and after
• Market
• Operation

Where to compete, how to position by segment and how your team executes, in pages that evolve with operational feedback.

• Strategy document
• Positioning
• Playbook

Every artifact is generated from the client folder, the same source that feeds the Workspace. No one copies numbers into slides.

• Interactive HTML
• Shareable