Clay?

NOT REALLYA prototype of the interface is possible, but the durable product depends on difficult systems beyond the UI. Build a safer adjacent workflow instead.

BUILD FIRST

A non-sensitive or read-only version of the core experience
A manual review step wherever correctness or trust matters
A narrow companion tool that removes a real piece of work

LEAVE FOR LATER

Regulated decisions, payments, or irreversible actions
Claims of correctness, compliance, or uptime that the MVP cannot support
Sensitive integrations until security, support, and auditability are ready

RESEARCH NOTES

What this tool actually does.

Checked 2026-09-06

Clay is a B2B go-to-market data and workflow platform, not a traditional sales-outreach sequencer. Users build spreadsheet-like tables that feed through Claygent (an AI web researcher that browses public sources for custom fields) and Sculptor (a waterfall engine that queries 150+ data providers in sequence until a verified email match is found, charging credits only on hits). It offers a 200+ vendor data/AI marketplace, email verification flagging catch-all/role inboxes, auto-refresh scheduling, and bidirectional CRM sync (HubSpot, Salesforce, Pipedrive, Close). It also personalizes outbound email copy using 1st- and 3rd-party data, routes leads by intent signals, and automates meeting prep and follow-ups. Native integrations include Hunter, Harmonic, LinkedIn, Google Maps, and GitHub; webhooks and an HTTP API cover gaps. Sold as 'build systems to grow revenue' for GTM teams.

SOURCES

RESEARCHED BUILD PROMPT

Give this to your coding agent.

Build a focused replacement for Clay (clay.com), a B2B go-to-market data workflow tool. Core workflow: user pastes/imports an account list into a table, adds columns specifying target titles and custom data fields; an AI web-research agent (like Claygent) browses company sites, LinkedIn, and public pages to extract those fields; a waterfall enrichment engine (like Sculptor) queries a user-ordered sequence of data providers until it finds a verified email for the target title, consuming credits only on successful matches; results land in a reviewed table the user can push to HubSpot/Salesforce or export as CSV. User-facing requirements: spreadsheet-style table editor with column formulas, a provider picker letting users order 10+ enrichment providers in a waterfall, per-row AI research runs with configurable prompts, verification flags for catch-all/role inboxes, and one-click exports/CRM pushes. Technical requirements: Node/TypeScript or Python backend; a credit-billing counter that debits only on provider hits; a provider adapter interface so new providers drop in; a queue system for long-running enrichment jobs; webhook/HTTP API for external tooling; OAuth for HubSpot and Salesforce sync. Implementation approach: build the table UI in React with a cell-formula engine; run the AI research agent server-side with Playwright/Puppeteer and a structured-extraction LLM; implement the waterfall as a deterministic state machine with per-provider skip/retry rules; use Postgres for persistence and Redis for job queuing. Explicitly do not copy Clay's branding, logo, marketing copy, or UI layout — design a clean, neutral interface with your own visual identity. Out of scope: building a 200-vendor marketplace, paid data contracts, a Chrome extension, email sending/sequencing features, and multi-tenant billing. Include a README covering setup (env vars, provider API keys), how the waterfall and credit counter work, and known limitations (AI research can be slow or inaccurate, provider rate limits, verification is heuristic).

Visit the official product page