Verified contacts, the evidence behind each one and a first touch written from that evidence, loaded into Lemlist or Instantly as a paused sequence with the source quoted beside every line. The infrastructure under outreach, run by autonomous AI agents; sending stays with your team.
A hiring signal, a funding event, a stack change or an intent topic cited in each first touch, with the source linked.
Every row loaded into a sequence has a verified work email; catch-all and risky addresses are held aside and labeled.
Drafts land in your outreach tool paused. A person reads, edits and presses send inside the tool you already use.
A batch of drafts into Lemlist, Instantly or the CRM waits for a person exactly as previewed; research never does.
The template has two merge fields and the reply rate shows it. Real personalization needs the open role, the round, the stack change or the post the buyer wrote, and researching that by hand takes ten minutes a contact.
A list exported from one tool, uploaded to another, sent to addresses nobody checked. Bounces climb, the domain’s reputation follows, and the good contacts in the same batch pay for the bad ones.
The evidence sits in a spreadsheet, the draft sits in the outreach tool, and the link between them is a rep’s memory. When the SDR changes, the reasoning behind every touch leaves with them.
For each person, the agents collect what is true right now: the roles their company posted, the round it raised, the technology it adopted, the topics it is researching and the posts the person published, dated and linked.
Work emails found through a waterfall of licensed sources and verified; direct phones found where a source has one. Risky and catch-all addresses are labeled and held, so a sequence loads clean.
A draft per contact that opens with the specific signal, cites its source and makes one ask, in the tone and length your workspace sets. Every line traces to a fact the agents can show you.
Drafts and contacts pushed into Lemlist or Instantly as a paused sequence, into HubSpot, Pipedrive or Attio as notes on the contact, or into a sheet, after one approval. Sending stays in your tool.
A saved list, a CRM segment, a CSV or a mission that finds the accounts first. The workspace ICP and suppression list apply automatically.
Every contact re-verified at send time, missing emails found through the waterfall, role accounts and suppressed companies removed.
Hiring, funding, stack, news, intent and the person’s own posts collected and dated for each row; rows with nothing worth citing are flagged.
A first touch per contact from the evidence, in your voice, with the citation beside each claim so a reviewer can check it in seconds.
Drafts shown with their evidence and verification status; edit the ones you want, drop the ones you do not, set the batch to approve.
Contacts and drafts pushed into Lemlist or Instantly as a paused sequence, or into the CRM as notes, after one approval. Your team presses send.
Fifty accounts that posted the roles your product relieves, the executive who owns the pain verified, a draft per contact that quotes the posting, loaded into Instantly paused for a Monday review.
Companies in your ICP that announced a round in the last 14 days, the CEO and the functional buyer found and verified, a draft that names the round and connects it to a concrete next step.
Contacts at companies actively researching your category, with the topic and the timeline, and a draft that speaks to the question they are already asking.
Every closed-lost opportunity from last year re-enriched: who is still there, what changed at the company, and a draft that references the change instead of the old deal.
| By hand | With AstroFabric | |
|---|---|---|
| Personalization | First name, company, a template | The signal, the source and one ask, per contact |
| Research | Ten minutes per contact in six tabs | Gathered and dated by agents, shown beside the draft |
| Contact quality | Whatever the export contained | Verified at send time; risky rows held aside |
| Loading | CSV upload and column mapping | Pushed into Lemlist or Instantly paused, after approval |
| Sending | Automated by a tool nobody watches | By your team, in your tool, after reading |
| Reasoning | In the rep’s head | Evidence and citations on every row, kept in the run history |
Fifty accounts researched, the buyer verified at each, a first touch written from the strongest signal with the citation, loaded into Lemlist paused.
Companies that raised in the last two weeks, matched to the ICP, the buying committee verified, drafts that name the round and the next step, staged in Instantly.
Contacts at companies surging on your category topics, the topic and timeline attached, drafts that answer the question being researched, notes written to the CRM.
Three of a hundred. The full playbook library ships in every workspace, and anything you can describe becomes a mission.
The agents write and load. Every message is sent by a person from your outreach tool, and no capability on the platform can send mail in your name.
Contacts are re-verified when the batch is built, so a sequence loads with current addresses and the held rows are labeled with the reason.
Customers, open opportunities, competitors and the workspace suppression list are removed before a batch is staged.
Signals and people data come from licensed providers. The agents never scrape member networks.
The layer that hands your outreach tool finished inputs: verified contacts, the dated evidence on each account and person, and a first touch written from that evidence with citations, loaded as a paused sequence or CRM notes after approval. Agents research and draft; your team sends.
No. Drafts are loaded into Lemlist or Instantly as a paused sequence, or attached to CRM contacts as notes, and a person reads and sends from that tool. No capability on the platform sends mail, and that boundary applies to the API, MCP server and CLI as well.
Open roles and hiring velocity, the latest funding round, technology adopted or replaced, buying-intent topics with their timeline, dated news, leadership changes and the contact’s own recent posts. Each claim in the draft carries its source, and rows without a citable signal are flagged.
Workspace settings hold your tone, length, sign-off and the claims you never make. A mission can adjust any of them for one batch in plain language, and every draft is editable in review before it loads, with its evidence and citations still attached to the row.
Work emails are found through a waterfall of licensed sources and checked by a verifier when the batch is built. Valid rows load; catch-all and risky addresses are held and labeled so you decide. A lookup that finds nothing costs nothing.
Yes. Upload a CSV, choose a saved list or pick a CRM segment. The agents verify each contact, gather the evidence, write the drafts and stage the load, and the original columns stay on every row. Rows outside the ICP or on the suppression list are flagged before drafting starts.
A verified contact is a few credits and each draft is a few credits, shown before the batch runs. Evidence gathering, suppression and loading into your tools are included, misses on contact finding are free, and plans start at $49 per month.
Pick the list. Verified contacts and evidence-backed drafts land in your outreach tool paused, ready for your team to send.