Prospect Identification
Targeted prospect lists built from your real ICP, enriched against live data, deduped against your CRM, and delivered ready to work. Your reps stop scrubbing rows and start running plays.
Overview
Most sales teams don't have a prospecting problem. They have a list problem. The database vendors will sell you 200 million contacts and call it pipeline fuel. What lands in your reps' hands is a CSV that's 30% out of date, 20% already in your CRM, 15% off-ICP, and 10% people who left the company two reorgs ago. The remaining quarter is the actual work, buried under the noise.
This agent fixes the upstream problem. It builds prospect lists from your live ICP definition: industry, geography, headcount, title, tech stack, funding signal, intent data, anything you can describe. It then pulls fresh data on every record, suppresses anyone already touched in your CRM or sequenced in your engagement tool, and delivers a clean working list in the format your team uses. Reps spend their hours selling. The list does the filtering before they ever see it.
Capabilities
- Builds targeted prospect lists from your defined ICP criteria across industry, geography, size, role, seniority, and buying signals
- Enriches every record with current firmographic, technographic, and contact-level data at the moment of list generation, not from stale cached files
- Suppresses contacts already in your CRM, already in active sequences, or recently worked by another rep, so no one gets double-touched
- Filters out departed employees, generic role accounts, and deliverability-risk addresses before the list reaches a rep
- Scores each prospect against your ICP fit model so reps know which rows to work first
- Delivers lists in the format and destination your team actually uses: CSV, CRM list, sequence enrollment, or direct push to your sales engagement platform
Example Output
Request: Territory list for Jordan Rivera, mid-market AE
Request: Territory list for Jordan Rivera, mid-market AE. ICP: B2B SaaS companies, 100-500 employees, headquartered in US Northeast, with a VP-level marketing or RevOps title, currently using a competing marketing automation platform, not currently in the CRM.
Agent output, delivered as a 247-row working list, scored and ranked:
List name: Jordan_NE_MidMarket_2026-05 Total records: 247 (filtered from an initial pool of 2,840) Suppression applied: 1,612 already in CRM, 681 in active sequences, 184 departed employees, 116 generic role accounts
Top 5 ranked records (full list attached):
- Priya Shah · VP Marketing · Lattice Health · Boston · 280 employees · Marketo + Salesforce · Series C, raised 8 weeks ago · Fit score: 94
- Marcus Bell · VP RevOps · Cohere Logistics · Brooklyn · 410 employees · HubSpot Enterprise · Hiring 3 RevOps roles · Fit score: 91
- Sarah Lin · Head of Marketing · Northwind Analytics · Cambridge · 165 employees · Pardot · Recent CMO change · Fit score: 89
- David Okafor · VP Demand Gen · Tessera AI · Manhattan · 220 employees · Marketo · Posted about MAP migration on LinkedIn · Fit score: 87
- Elena Vasquez · VP Marketing · Helix Bio · Stamford · 340 employees · Eloqua · Series B extension last quarter · Fit score: 86
Logged: List delivered to Jordan Rivera's Outreach instance. Audit trail written to CRM. ICP definition snapshot saved for next refresh in 30 days.
Agent Workflow
Load the ICP definition
The agent pulls your ICP rules from a versioned definition you control. ICP is treated as code, not as tribal knowledge. Changes are tracked, so a list built today can be reproduced or compared against one built last quarter.
Source the candidate pool
The agent queries connected data sources: intent data feeds, technographic vendors, public signal sources etc. and assembles a candidate pool that matches the ICP criteria. The pool is intentionally wider than the final list. Filtering happens next.
Enrich every record
The agent pulls fresh data on every candidate at runtime: current title, current employer, current contact info, current company headcount, current funding state, current tech stack. Stale records are flagged or dropped.
Suppress conflicts and duplicates
The agent compares the candidate pool against your CRM, your sequence tool, and recent rep activity. Anyone already in motion gets removed. Anyone touched in the last 90 days by another rep gets removed. The list that survives is genuinely net-new work.
Filter for deliverability and signal quality
Departed employees, generic role accounts, role-based emails, hard-bounce-risk addresses, and known opt-outs get filtered before a rep ever sees them. Domain health stays intact. Compliance stays intact.
Score and rank
The agent scores each surviving record against your ICP fit model, firmographic match, role match, intent signals, recency of buying triggers, and ranks the list so reps work the strongest rows first. The top of the list is where pipeline lives.
Deliver in the working format
The agent pushes the final list to where the rep actually works: CSV download, CRM list, Outreach or Salesloft sequence enrollment, or a direct task queue. No format friction, no copy-paste, no "I'll get to it after I clean it."
Log, version, and schedule the refresh
Every list is logged with its ICP definition snapshot, suppression rules applied, source providers used, and timestamp. Lists can be refreshed on a cadence, weekly, monthly, on a trigger, so the territory stays alive without anyone re-briefing the agent.