Skip to content
EarthNWe Labs
LiveSaaS

Calamus LeadX

Agents do the research.
You approve the send.

LeadX runs the part of outreach that does not scale: finding the right people, reading up on their business, and writing something worth replying to. Every message waits for a human before it goes out.

labs.earthnwe.com/calamus/leadx
The Calamus LeadX Approval Studio. The left column shows the prospect, a fit score, and the research the agents found about their company. The right column shows the drafted email built from that research, with Regenerate and Approve and send actions. Prospect and message shown are illustrative.

Performance at Scale

Agents run the parts of outreach that don’t scale. Results speak for themselves.

A draft moves from the writer to the critic. Scoring at least 70 sends it on to you; below that it returns to the writer, and only twice before it stops for a person.
5
Agents in the crew
3
Enrichment sources
15s
Reply triage
0
Sends without you

Powered by

Where the data comes from.

Enrichment and deliverability, before a word gets written.

  • ApolloCompany and contact data

    The first pass: who works there, what the company does, and whether the shape of it matches the segment you described.

  • Hunter.ioEmail discovery

    Finds the address, and says how confident it is. A guess is marked as a guess rather than sent.

  • ZeroBounceAddress verification

    Checks the address is real before anything is drafted for it, so the sender reputation is not spent on bounces.

  • GeminiDrafting and reply reading

    Drafts the message from the research, and reads what comes back well enough to tell a real reply from an auto-responder.

Five agents, one hand-off chain.

Each agent does one job and passes the result on. The chain stops at the approval gate — that stop is the product, not a limitation.

  1. ProspectorScored for fit
  2. Researcher3 sources
  3. WriterOne finding
  4. CriticScore >= 70
  5. YouApproval gateWaits for you
  6. ManagerReply triage
01

Prospector finds them

Describe the segment in plain language and the agents go looking. Results come back scored for fit, not just matched on a filter.

02

Researcher reads up

Enrichment across Apollo, Hunter.io and ZeroBounce, plus what the company has said publicly — an opening, an announcement, a change in how they buy.

03

Writer drafts it

One message built on that specific finding. The research sits next to the draft so you can see exactly what the claim rests on.

04

Critic scores it

Every draft is checked for hallucinations, flattery, and filler. If the score is under 70, it is rejected and rewritten automatically.

05

You approve it

Nothing sends on its own. Read it, refine it in a sentence, regenerate, or send. Verified recipient required.

06

Manager works the replies

Threads are polled continuously. The manager agent reads sentiment, drafts the follow-up, marks the hot ones and retires the dead ones.

Where the work actually happens.

Six screens. The second one is where you will spend your time.

labs.earthnwe.com/calamus/leadx/command-centerDetail
A partial view of the Command Center screen in LeadX. Figures shown are illustrative.

Part of one screen · open the tool to see the rest

Get in touch

See a draft before you trust it.

The fastest way to judge LeadX is to read one message it wrote and the research it was built on. Open the studio, or talk to us about the setup.