← Back to selected work

Early Jev AI use case

Lead Cleanup

An early Jev AI use case that turns messy contact exports into reviewed, outreach-ready leads while keeping decisions explainable.

Jev AICSV and XLSXPolicy rulesReview queue

On Jev AI's second day after launch, I identified lead qualification as a practical use case and integrated it into an existing workflow. Lead Cleanup accepts CSV or XLSX contact exports, applies clear rules, finds duplicates, and sends uncertain cases for human review.

RoleProduct and implementation
Business questionWhich contacts fit our outreach criteria, and which need a person to decide?
ArchitectureVite browser app, Node and Netlify Functions, deterministic rules, server-side Jev suggestions

THE OPERATING FLOW

01Upload and map a contact export without changing its source rows
02Apply versioned rules, duplicate checks, and company-level checks
03Review uncertain leads and export ready contacts with an audit file

CASE STUDY

The challenge

Contact exports contain inconsistent job titles, duplicates, and old qualification labels that cannot be trusted as final decisions. Sales teams need to know which leads are ready for outreach and why.

What I designed and built

  • Kept explicit policy rules and duplicate handling deterministic so the same input gets the same decision.
  • Integrated Jev AI quickly into the existing flow for typed suggestions on ambiguous seniority and function; people approve or reject those suggestions.
  • Separated the ready-to-contact export from the audit and review queue so provisional leads are not sent out as approved.

Outcome

  • A concrete early Jev AI use case that fits into an existing, explainable lead workflow.
  • A review queue and audit file that show why each contact was accepted, rejected, or held.

Private contact exports and credentials are not published. AI suggestions remain provisional until reviewed.