An end-to-end scripted prospecting pipeline
Manual prospecting didn't scale. The bottleneck sat in data quality.
Context
Suris Code is an Argentine software factory. I was their Sales & Growth Lead from August 2025 to July 2026, with the whole funnel on me.
Manual prospecting didn't scale, and the problem was data quality. Every batch had to be extracted, filtered against the ICP, checked for deliverable addresses and normalised before it went anywhere near an outbound sequence. Sending volume gets fixed with more inboxes. This doesn't.
What I built
An end-to-end pipeline, scripted with Claude Code:
SalesQL → ICP filtering → email matching → deliverability verification
→ my own SQL database (Contact Core) → Smartlead via API
Pull contacts, drop anything that isn't ICP or lacks a usable, correctly prioritised email, verify deliverability, normalise first names and company names into text that reads like a human wrote it, load into Smartlead as a clean CSV.
What happened
One batch of verified, sequence-ready prospects used to take 80 minutes of manual work.
| Step | Before |
|---|---|
| Extraction and ICP filtering | ~30 min |
| Email verification and list cleanup | ~20 min |
| Name and company normalisation | ~15 min |
| CSV build and Smartlead import | ~15 min |
Now it takes 20 minutes, with a manual review step inside that. 75% less time per batch.
The pipeline runs roughly 2,000 prospects per week and yields about 1,400 clean, usable contacts after filtering. The ceiling comes from inbox sending capacity and the rest of the team's workload.
The iteration
Email verification was the slowest link in the chain. CaptainVerify flagged around 20% of incoming contacts as invalid, doing its job well, but each run took about 20 minutes and cut the flow between steps.
I moved verification to Bouncer, over API, with results in near real time. The wait went away and the data-quality bar stayed.
Limits
The scripting runs on Claude Code: comfort with tools and integrations.
I never measured the effect on deliverability. I have the time saved and the volume coming through; how much delivery improved after Bouncer, I don't.