Deepline for GTM Engineers
Agents write the pipeline, Deepline runs it. One CLI for waterfall enrichment across 76 integrations, AI scoring, email validation, and sequencer push. Built for Claude Code, Codex, and Cursor.
Audience
Who this is for
GTM engineers are the fastest-growing non-technical user group for AI coding agents. If you live in Claude Code, Cursor, or Codex and need reliable, programmable enrichment without spreadsheet-based tools, Deepline is the missing layer.
- You write scripts, not formulas. Your workflow belongs in a git repo, not a SaaS UI.
- You need agents to run enrichment autonomously, not click through a browser.
- You have provider API keys (Apollo, Crustdata, PDL) and want to use them directly.
- You process thousands of rows and can't afford arbitrary table caps or auto-run credit burn.
Workflows
5 workflows GTM engineers run with Deepline
Step 01: Waterfall lead enrichment
200 badge scans from a conference. Waterfall across Apollo and PDL for emails, validate with LeadMagic, output a clean list. One command, high email coverage. No formula columns.
deepline enrich --input badge-scans.csv --output enriched.csv \
--with-waterfall "email" \
--with 'apollo=apollo_people_match:{"first_name":"{{First Name}}","last_name":"{{Last Name}}","organization_name":"{{Company}}"}' \
--with 'pdl=peopledatalabs_enrich_contact:{"first_name":"{{First Name}}","last_name":"{{Last Name}}","company":"{{Company}}"}' \
--end-waterfall \
--with 'verify=leadmagic_email_validation:{"email":"{{email}}"}'Step 02: Email verification and list hygiene
Validate every email before it hits a sequencer. LeadMagic checks deliverability, catches disposable domains, and flags risky addresses. Keep bounce rates under 2% and protect sender reputation.
deepline enrich --input contacts.csv --output verified.csv \
--with 'verify=leadmagic_email_validation:{"email":"{{Email}}"}' \
--with 'filter=run_javascript:"return row.verify_status === \"valid\" ? row : null"'Step 03: ICP scoring with AI
Enrich leads with company data, then use deeplineagent to score each one against your ICP criteria. AI runs deterministically per row. No batch degradation where everyone gets scored "A" at scale. JSON mode returns structured scores you can filter on.
deepline enrich --input leads.csv --output scored.csv \
--with 'company=crustdata_companydb_autocomplete:{"field":"company_name","query":"{{Company}}"}' \
--with 'score=deeplineagent:{"prompt":"Score this lead 1-10 against our ICP (Series B+ SaaS, 50-500 employees, NA/EU). Company: {{company}}. Title: {{Title}}. Return JSON with score, tier (A/B/C), and reason.","json_mode":true}' \
--rows 0:1 # pilot on 2 rows firstStep 04: Build a TAM list from scratch
Start with target companies, find decision-maker contacts via Apollo, waterfall emails, validate, and score, all in one pipeline. The agent handles the entire workflow from company list to scored, verified contacts.
# Build company list + find contacts + enrich + validate
deepline enrich --input target-companies.csv --output tam.csv \
--with 'company=crustdata_companydb_autocomplete:{"field":"company_name","query":"{{Company}}"}' \
--with 'contacts=apollo_org_search:{"organization_name":"{{Company}}","person_titles":["VP Sales","Head of Revenue"]}' \
--with-waterfall "email" \
--with 'apollo_email=apollo_people_match:{"first_name":"{{contacts.first_name}}","last_name":"{{contacts.last_name}}"}' \
--with 'pdl_email=peopledatalabs_enrich_contact:{"first_name":"{{contacts.first_name}}","last_name":"{{contacts.last_name}}"}' \
--end-waterfall \
--with 'verify=leadmagic_email_validation:{"email":"{{email}}"}' \
--with 'score=deeplineagent:{"prompt":"Score 1-10 for ICP fit: {{company}} {{contacts}}","json_mode":true}'Step 05: Sequence automation
Push enriched, verified, scored contacts directly to Instantly, Lemlist, HeyReach, or Smartlead. No manual CSV exports, no copy-paste. The agent enriches, scores, filters, and sequences in one session.
# After enrichment, push top-scoring leads to sequencer deepline sequence --provider instantly --campaign tam-q1 --input tam.csv # Or push to multiple sequencers deepline sequence --provider lemlist --campaign outbound-q1 --input tam.csv deepline sequence --provider heyreach --campaign linkedin-q1 --input tam.csv
Step 06: Slack approval workflow
Deploy workflows that send you a Slack message with results, let you edit and approve, then continue to the next step, all in one flow.
Claude Code deploys a workflow with Slack bot integration:
deepline workflows deploy lead-review \ --trigger webhook \ --notify slack \ --await-approval \ --on-approve "deepline sequence --provider instantly --campaign q2"
Agent Orchestration
Agents write the pipeline, Deepline runs it
Deepline ships a Claude Code skill with slash commands for every enrichment workflow. Describe what you want in plain English. The agent reads the tool catalog, writes the pipeline, runs it, and reviews the output. Context from each step carries forward into the next.
Example: just tell Claude Code what you need
> "I have a CSV of 200 leads from a conference. Enrich emails > with waterfall, validate them, score against our ICP, and > push the top 50 to Instantly." # Claude Code reads the Deepline skill and runs: deepline enrich --input conference.csv --output enriched.csv \ --with-waterfall "email" \ --with 'apollo=apollo_people_match:...' \ --with 'pdl=peopledatalabs_enrich_contact:...' \ --end-waterfall \ --with 'verify=leadmagic_email_validation:...' \ --with 'score=deeplineagent:...' \ --rows 0:1 # pilot first # Reviews results, then runs full batch + sequences: deepline sequence --provider instantly --campaign conference-q1 --input enriched.csv
No copy-pasting commands. No reading docs. Works the same way with Codex and Cursor. Any agent that can execute shell commands can orchestrate Deepline.
CLI vs Spreadsheet
Why the CLI beats the spreadsheet
| Advantage | Details |
|---|---|
| Reproducible | Your pipeline is a shell command. Check it into git. Run it next quarter. Share it with your team. No one can deprecate your workflow. |
| Version-controlled | Enrichment logic lives in scripts alongside your code. Diff it, review it, roll it back. Not trapped in a SaaS UI. |
| Agent-orchestratable | Claude Code, Codex, and Cursor call deepline enrich the same way they call git or curl. No browser automation, no MCP overhead — 5-10x fewer tokens per operation. |
| No row limits | Process 500K rows with the same command as 500. The limit is your disk space, not an arbitrary table cap. Pilot with --rows 0:1 before committing. |
Configurations
Common GTM engineer setups
| Setup | What you need | Cost |
|---|---|---|
| BYOK Starter | Apollo API key (free tier: 10K credits/mo), Crustdata API key, LeadMagic API key (email validation) | Free — you pay providers directly |
| Managed Credits | Deepline account with credits. All 76 integrations available instantly. Per-row billing, no minimums. | Pay per enrichment, monthly billing caps |
| Full Waterfall + Validation | Apollo + PDL + Crustdata (waterfall), LeadMagic (email validation), Instantly or Smartlead (sequencing), deeplineagent for ICP scoring | Highest coverage, pay per provider hit |
Trusted by GTM teams
Frequently Asked Questions
1Do I need to leave my terminal to use Deepline?+
No. Deepline is a CLI tool. You install it with curl, configure API keys, and run enrichment commands directly from your terminal. There is a browser-based playground for inspecting results, but it runs locally and is optional.
2How does Deepline work with Claude Code?+
Deepline ships a Claude Code skill with slash commands for every workflow. When you describe what you want in Claude Code — like 'enrich this CSV with waterfall email lookup' — the agent reads the skill, writes the deepline enrich command, runs it, and reviews the output. No copy-pasting or manual setup required.
3Can I use my own Apollo, PDL, or Crustdata API keys?+
Yes. Deepline supports bring-your-own-key (BYOK) for all 76 integrations. Connect your existing provider accounts and use Deepline for free. You only pay what you already pay your providers. Alternatively, use Deepline-managed credits if you prefer not to manage keys.
4What is waterfall enrichment and why do GTM engineers need it?+
Waterfall enrichment tries multiple data providers in sequence and stops at the first match. For example, try Apollo for an email, fall back to PDL if Apollo misses, then validate with LeadMagic. In Deepline, this is a single --with-waterfall flag. In spreadsheet tools, it requires dozens of formula columns and manual conditional logic.
5How does Deepline pricing compare to Clay for GTM engineers?+
With BYOK, Deepline is free — you pay your providers directly. With managed credits, you pay per enrichment with no minimums. There are no auto-run surprises because execution is always explicit. Monthly billing caps prevent accidental overspend. Clay charges $495+/mo for Growth with credit bundles that can auto-burn.
6Does Deepline work with Codex and Cursor too?+
Yes. Deepline is a standard CLI tool — any AI agent that can execute shell commands can use it. Claude Code, OpenAI Codex, and Cursor all call deepline enrich the same way they call git or curl. No browser automation, no MCP setup, no special integration required.
Try Deepline in 30 seconds
Install the CLI and run your first enrichment. Free with your own API keys.