Rank LinkedIn outbound leads and pick the opener for each
Jev scores each HeyReach lead, each first-message variant and the chance of a reply, then writes draft campaigns for a person to approve.
DifficultyIntermediate
Setup time1-2 hours
CategoryDecision models (Jev)
The use case
Sales teams running LinkedIn outreach in HeyReach send the same opener to long lists with mixed fit. Jev scores every lead against buying signals, scores each written message variant for that person, and estimates reply chance. The tool writes ranked lead lists and draft campaigns back to HeyReach; it does not start campaigns or send messages. Once reply history exists, it reranks using what earned replies. A rep reviews each draft campaign and the lead list before anything is sent.
Example workflow
- Connect HeyReach and TypeSafe API keys
- Jev scores leads in each list against your signals
- Jev scores each opener variant per lead and estimates reply chance
- Tool writes ranked lists and draft campaigns back to HeyReach
- Rep reviews drafts and launches campaigns by hand
Tools
- Jev (TypeSafe AI)
- HeyReach
- Claude Code
Risks and guardrails
- Poor fit signals push the wrong leads to the top
- Messaging people at scale can breach LinkedIn rules
- Reply-based learning needs enough history to mean anything
Sources
Written by Agents Lookup. Credits and official pages for the tools this use case names:
- GitHub repo heyreach-jev-bot by matthew004-web: github.com/matthew004-web/heyreach-jev-bot
- HeyReach (official page): www.heyreach.io (checked 2026-11-09)