Choose a surface
Decide between the Agent, People Sets, synchronous APIs, CLI, and MCP.
Agent API
Choose POST /v1/agent when the input is a goal and the work may require multiple searches, source selection, judgment, and synthesis.
Good examples:
- Research a seed account and return stage-matched lookalikes.
- Verify a person, map likely buyers, and draft evidence-grounded outreach.
- Analyze supplied account data and add current public signals.
People Sets
Choose People Sets when the deliverable is a persistent, criteria-verified dataset. Sets can run asynchronously, emit progress, accept enrichments, and export results.
Synchronous search
Choose POST /v1/search for a fast one-shot list of up to 100 people. It is useful for interactive autocomplete, small cohorts, and low-latency agent tools.
Enrichment
Choose the enrichment APIs when you already know the identity and need selected profile, email, phone, or social fields.
CLI
Choose the CLI for terminal exploration, local scripts, manual QA, and smoke tests.
MCP
Choose the hosted MCP server when people research should
be available as native tools inside ChatGPT, Claude, or another MCP client. Add
https://api.ploid.com/mcp; Ploid handles sign-in and scoped workspace access.
A useful rule
Use the highest-level surface that naturally matches the user's request. Reach for a lower-level endpoint only when your application needs tighter control over the workflow.