New batch

Upload, configure operations, name it, review. The batch lands in `pending` until the worker picks it up (worker is the next milestone — for now leads sit unclassified).

File

CSV or TXT. First row must be a header. Max 100MB.

Drop a CSV or TXT here

or choose a file

.csv (comma, tab, semicolon) .txt (one email per line)

Recognized CSV columns: email · domain · company · full_name. Unrecognized columns are ignored.

Operations

Pick what the workers should do to each row. Combine freely — they run in this order.

Full pipeline (P1 → P1.5 SPF → P2 PTR → P3 HTTP → P4 region)

Best recall. ~10–100 leads/sec/worker. Recommended unless you're pre-filtering.

MX-only classification

Just P1. ~50k leads/sec/worker. Use as a fast pre-filter; re-run later with full pipeline.

Detect duplicates against existing batches

Cross-reference the upload against every prior batch. Always recommended.

Skip P3 HTTP if earlier phases produced nothing

Saves the most expensive step on cold leads. Trade-off: ~10% lower recall for ~50% throughput gain.

Enrich industry (Stage 2 firmographic)

After provider detection, run the Tier 0 rules over each lead's network footprint to tag industry. Shopify auto-resolves; misses go to review. Near-zero cost — reuses Stage 1 DNS.

Name & metadata

Helps you find it later. Tags and notes are optional.

Batch name

Required. Will appear on cards in the Batches view.

Review

Preview is computed client-side. The server runs cross-batch dedup at upload time.

Pick a file to see a preview.

Cancel