Agent-callable GTM data layer. Enrich, validate, route.

Overview

What is the GTM data layer?

The GTM data layer is the orchestration and execution system between your data sources (Apollo, PDL, Clearbit, Crustdata) and your destinations (HubSpot, Salesforce, Instantly, Lemlist). It handles enrichment, validation, scoring, routing, and activation—not as point-and-click workflows in a UI, but as typed, versioned Plays that coding agents can inspect, run, and own.

Deepline is the agent-callable GTM data layer. 106+ integrations accessible through CLI, SDK, and MCP for Claude Code, Cursor, Codex, and any shell-capable coding agent.

Capabilities

B2B enrichment workflows by category

Waterfall

Multi-provider waterfall enrichment

Deepline waterfalls query providers in sequence and stop when one returns valid data. If Apollo misses, the Play tries the next declared stage. You pay only the provider that succeeds.

Common waterfall providers for work emails:

Maintained waterfall Plays are discoverable, inspectable, and versioned. The Play contract declares provider stages, validation rules, and fallback behavior. Run a pilot on representative data before scaling.

# Search the live Play catalog
deepline plays search "verified work email waterfall" --json

# Inspect the contract
deepline plays describe prebuilt/name-and-domain-to-email-waterfall-batch --json

# Run a pilot
deepline plays run prebuilt/name-and-domain-to-email-waterfall-batch --csv leads.csv --watch

Database

Built-in PostgreSQL database

Every Deepline workspace includes a PostgreSQL database with auto-resolved tables:

• dl_resolved.contacts — Deduplicated contact records with enrichment history
• dl_resolved.accounts — Company records with firmographics and signals
• dl_resolved.events — Enrichment event log with provider receipts and usage

Query your GTM data with SQL. Export anytime. Build custom dashboards, scoring models, or applications on top of enriched data without exporting to a separate warehouse.

-- Find contacts enriched in the last 7 days
SELECT * FROM dl_resolved.contacts
WHERE enriched_at > NOW() - INTERVAL '7 days';

-- Check provider coverage by source
SELECT tool_id, COUNT(*), AVG(cost_credits)
FROM dl_resolved.events
GROUP BY tool_id;

Agent workflows

How coding agents use the GTM data layer

Tell Claude Code, Cursor, or Codex what you need. The agent uses the installed Deepline skill to discover, inspect, and run workflows:

Example: Enrich inbound leads

"Enrich these inbound leads with company size, industry, and verified work email.
Validate emails before pushing to our Instantly campaign."

The agent:

1. Searches: deepline plays search "company firmographic enrichment email waterfall" --json
2. Inspects: deepline plays describe <play-name> --json
3. Runs a pilot: deepline plays run <play-name> --csv leads-pilot.csv --watch
4. Reviews output, provenance, usage
5. Scales: deepline plays run <play-name> --csv leads-full.csv --watch
6. Validates: deepline tools execute leadmagic_email_validation ...
7. Pushes: deepline tools execute instantly_add_to_campaign ...

Example: Build an outbound campaign

"Build an outbound campaign targeting VPs of Engineering at Series B SaaS companies.
Find 200 contacts with verified emails."

The agent:

1. Finds companies matching the ICP (Apollo company search)
2. Discovers decision-makers at each company (Apollo people search)
3. Waterfalls for verified work emails (Hunter, Prospeo, Leadmagic, PDL)
4. Validates deliverability (Leadmagic)
5. Pushes to sequencer (Instantly, Lemlist, HeyReach)

One prompt, full pipeline. The agent reads the Deepline skill and builds the workflow automatically.

Pricing

BYOK or managed

• BYOK mode: Free. Connect your own provider API keys and pay providers directly. Deepline charges no platform fee.
• Managed credits: Transparent per-operation pricing. Check the tool contract before running: deepline tools describe <tool-id> --json

Inspect billing first. Every tool shows its pricing unit before execution. Set spending caps from the dashboard to control costs.

Related features

Learn more

FAQ

Common questions

Get started

Install Deepline

npm install -g deepline@latest
deepline setup --json
deepline plays search "enrichment" --json

For Claude Code and Cursor: code.deepline.com/INSTALL.md