Integration

Apollo + Claude Code Integration via Deepline

Use Apollo's People Match API from a checked Deepline Play. Search the live catalog for Apollo people match, describe the selected contract, and run it against a pilot CSV.

Direct answer

If your query is "Apollo account record company name verification"

The primitive answer is simple: Apollo matching works best when you give Claude Code an explicit person or company key set, then let Deepline send the structured payload to Apollo. For contact matching, the strongest inputs are first_name, last_name, organization_name, domain, and optionally email. For company-oriented verification, the strongest keys are the canonical company name and domain before you ask the agent to enrich downstream CRM or sequencer records.

What this means in practice:

• If you know the company name and domain, normalize those first, then ask Apollo for the matching person or organization record.
• If you only have a fuzzy company label, do not treat Apollo as a truth engine by itself. Use the domain and company name together, or route through a Deepline waterfall.
• If your workflow depends on whether the company record is actually the right account, log the exact input fields and the returned organization fields so the agent can explain why the match was accepted.

Primitives

Match inputs, outputs, and failure cases

Overview

What This Integration Does

Deepline connects your Apollo account to Claude Code via a native CLI skill. Once connected, Claude Code can enrich contacts via Apollo's People Match endpoint, combine Apollo with other providers in a waterfall, and push results to sequencers, all from slash commands or plain English.

No browser automation. Connect however your agent prefers: the Deepline MCP server for typed tool calls, or the CLI, which Claude Code calls directly the same way it calls git or curl. Both run the same waterfall logic against the same providers.

In Claude Code, ask for Apollo people matching on your CSV. The agent searches and describes the live Play catalog, checks owned source when customization is needed, and runs the selected Play with --watch.

Quick Start

Quick Start

Three steps to enrich contacts with Apollo from Claude Code:

Step 01: Install Deepline CLI

npm install -g deepline@latest && npm exec --yes --package=deepline@latest -- deepline setup --json

Step 02: Connect your Apollo API key

Open the Deepline dashboard integrations settings and add your personal Apollo API key.

Step 03: Run your first enrichment

deepline tools search "Apollo people match" --json
deepline tools describe apollo_people_match --json

# Claude Code authors a CSV Play from the inspected Apollo contract.

deepline plays check apollo-people-match.play.ts
head -n 4 leads.csv > leads-pilot.csv
deepline plays run --file apollo-people-match.play.ts --csv leads-pilot.csv --watch

Step 04: Inspect the company match fields before writing anywhere else

deepline plays run --file apollo-people-match.play.ts --csv leads.csv --run-id-file apollo-match-run.json --watch
RUN_ID=$(jq -er '.runId' apollo-match-run.json)
deepline runs export "$RUN_ID" --dataset result.rows --out apollo-match-review.csv

This gives you a review file before you push anything into Salesforce, HubSpot, or a sequencer. For account-verification use cases, that inspection step is where you confirm the company/domain pair is real before you automate the next action.

Why Deepline

Why Use Deepline vs Calling Apollo's API Directly

You could write custom code to call Apollo's People Match endpoint. Here is why teams use Deepline instead:

Data Fields

What Apollo Data You Can Access via Deepline

Deepline exposes Apollo's full People Match API. Every data point Apollo returns is available as a column in your output CSV:

API Details

Apollo People Match Endpoint

Under the hood, Deepline calls Apollo's POST /v1/people/match endpoint. Here is the request structure Deepline constructs from your CSV columns:

// Apollo People Match - what Deepline sends per row
POST https://api.apollo.io/v1/people/match
{
"first_name": "Jane",
"last_name": "Doe",
"organization_name": "Acme Corp",
"email": "jane@acme.com" // optional - improves match rate
}

// Response includes:
// email, phone, title, seniority, linkedin_url,
// organization.name, organization.industry,
// organization.estimated_num_employees, ...

You never write this code. Deepline maps your CSV headers to the request fields and flattens the response into CSV columns automatically.

Waterfall

Apollo in a Waterfall

Apollo coverage varies by ICP, region, field, and input quality. If representative testing shows gaps, inspect the current provider contracts and keep the approved fallback order in one owned Play:

deepline tools search "Apollo Crustdata PDL work email" --json

# Claude Code authors the fallback order from the inspected live contracts.

deepline plays check apollo-fallback-email.play.ts
deepline plays run --file apollo-fallback-email.play.ts --csv leads.csv --watch

The Play stops when a stage returns an accepted result. Later stages do not execute after that point. Billing still depends on the pricing unit of every stage that ran. Learn more about waterfall enrichment.

For a provider-level reference, see Apollo in the GTM Stack directory, Crustdata in the GTM Stack directory, and the Apollo to Instantly Claude Code workflow.

Natural Language

Using Apollo with Claude Code in Plain English

You don't need to memorize CLI flags. After installing Deepline, just tell Claude Code what you want in plain English. It reads the skill docs and calls the right Deepline commands automatically.

You say: "Enrich these leads with Apollo"

deepline tools describe apollo_people_match --json
deepline plays check apollo-people-match.play.ts
deepline plays run --file apollo-people-match.play.ts --csv leads.csv --watch

You say: "Find emails for these companies using Apollo and fall back to PDL"

deepline tools search "Apollo PDL work email" --json
deepline plays check apollo-pdl-email.play.ts
deepline plays run --file apollo-pdl-email.play.ts --csv leads.csv --watch

You say: "Enrich this list with Apollo, validate the emails, and score against our ICP"

deepline tools search "Apollo email validation ICP scoring" --json
deepline plays check apollo-qualified-leads.play.ts
deepline plays run --file apollo-qualified-leads.play.ts --csv leads.csv --watch