Sales Intelligence MCP — B2B Lead Gen for AI Agents

One MCP endpoint. Ten lead-gen + enrichment tools. Drop it into Claude Desktop, Cursor, Cline, or any agent that speaks the Model Context Protocol — and your AI assistant suddenly has the prospecting power of Apollo, ZoomInfo, Clearbit, and Lusha combined, billed per call instead of per seat.

This MCP server bundles 10 production NexGenData actors behind a single streamable-HTTP MCP transport. Your agent calls find_company_emails("stripe.com") or find_b2b_leads(industry="fintech", country="Singapore") and gets clean JSON back. No API keys to juggle, no rate-limit headaches, no $1,000/month Apollo seat that one analyst uses twice a week.

Who This Is For

• AI-powered SDR tooling — give your sales agent the ability to research, qualify, and enrich prospects autonomously.
• Founders building lead-gen automations — replace a four-vendor sales stack (Apollo + ZoomInfo + Wappalyzer + Crunchbase Pro) with one MCP server.
• Sales engineers + RevOps — drop into Cursor and let Claude write enrichment loops over your CRM exports.
• AI agent frameworks — LangChain, Autogen, CrewAI, Smolagents users who want sales-intelligence tools that just work.

The 10 Tools

Tool	What it does	Upstream actor
find_company_emails(domain)	Discover role-based + verified business emails for a domain	nexgendata/company-email-finder
enrich_company(domain_or_name)	Enrich one company with domain, industry, employees, socials, emails	nexgendata/company-enrichment-tool
aggregate_company_profile(name, include_funding=True)	Full company profile aggregated from homepage + /about + /careers + Crunchbase + JSON-LD	nexgendata/company-data-aggregator
find_b2b_leads(industry, country, employee_band, limit)	Discover B2B leads matching industry / geo / size filters	nexgendata/b2b-leads-finder
enrich_lead_list(domains)	Bulk-enrich a list of domains with contact + social data	nexgendata/lead-list-enricher
detect_hiring_signal(company_slugs)	Detect hiring momentum (open roles, growth keywords, scaling signals) for target accounts	nexgendata/hiring-signal-detector
track_startup_funding(stage, sector, days_back)	Track recent funding rounds filtered by stage / sector / lookback window	nexgendata/startup-funding-tracker
detect_tech_stack(domain)	Detect website tech stack — frameworks, CMS, analytics, payment, hosting, marketing tools (Wappalyzer replacement)	nexgendata/wappalyzer-replacement
find_yc_companies(batch, stage)	Browse Y Combinator directory by batch (W26, S25, …) and status	nexgendata/yc-companies-directory-scraper
search_linkedin_jobs(query, location, posted_within_days)	Search LinkedIn for public job postings with title + company + posted date + description	nexgendata/linkedin-jobs-scraper

Every tool returns structured JSON wrapped in MCP's structuredContent envelope so agents can reason about it without an additional parsing step.

Connecting From Your MCP Client

The server runs in Apify Standby mode at:

https://nexgendata--sales-intelligence-mcp.apify.actor/mcp

You need an Apify token (free signup — credits included). Grab it from Apify Console → Settings → Integrations.

Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "sales-intelligence": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://nexgendata--sales-intelligence-mcp.apify.actor/mcp",
        "--header",
        "Authorization: Bearer YOUR_APIFY_TOKEN"
      ]
    }
  }
}

Cursor (~/.cursor/mcp.json)

{
  "mcpServers": {
    "sales-intelligence": {
      "url": "https://nexgendata--sales-intelligence-mcp.apify.actor/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Cline / Windsurf / Continue.dev

Use the same streamable-http transport. Point the client at the URL above with an Authorization: Bearer … header. Cline's MCP settings UI takes the URL and headers verbatim.

Python (raw FastMCP client)

from fastmcp import Client

async def main():
    headers = {"Authorization": "Bearer YOUR_APIFY_TOKEN"}
    async with Client(
        "https://nexgendata--sales-intelligence-mcp.apify.actor/mcp",
        headers=headers,
    ) as c:
        result = await c.call_tool("find_company_emails", {"domain": "stripe.com"})
        print(result.structured_content)

Example Agent Workflows

Workflow 1 — "Find 10 fintech companies in Singapore with hiring momentum and verified emails"

1. find_b2b_leads(industry="fintech", country="Singapore", limit=10)
       -> list of 10 fintech company domains in SG
2. detect_hiring_signal(company_slugs=[...10 slugs...])
       -> filter to ones with open roles or growth signals
3. enrich_lead_list(domains=[...filtered...])
       -> add verified emails, social profiles, contact pages

One natural-language prompt to your agent, three MCP tool calls, ~$1.20 in PPE charges, a contact-ready list of qualified leads. The equivalent workflow on Apollo + LinkedIn Sales Navigator is a $400/month minimum spend and 30+ minutes of UI clicking.

Workflow 2 — "Enrich this list of 100 domains with contact + funding data"

1. enrich_lead_list(domains=[…100 domains…])
       -> emails + social profiles + contact pages
2. for batch of 10:
       aggregate_company_profile(name, include_funding=True)
       -> industry + HQ + employee band + funding history + key people

Total cost: ~$11 for the full enrichment of 100 companies. ZoomInfo lookup tier comparable cost: $1,500+/month seat minimum.

Workflow 3 — "Track YC W26 batch + their tech stacks"

1. find_yc_companies(batch="W26")
       -> 50 freshly funded YC companies
2. for each company:
       detect_tech_stack(company.website)
       -> framework, CMS, analytics, payment, hosting

Perfect for devtool / DevRel teams who need to know what stack the latest cohort is building on. Total cost: ~$2.55 for 50 stack detections.

Workflow 4 — "Find AI startups that raised in the last 30 days and are hiring engineers"

1. track_startup_funding(sector="ai", days_back=30)
       -> list of recently funded AI startups
2. detect_hiring_signal(company_slugs=[...names...])
       -> filter to ones actively hiring engineers
3. find_company_emails(domain=...)
       -> for the top 20, fetch outreach addresses

Pricing — Pay Only For What You Call

We use Apify's pay-per-event (PPE) billing. Three events, transparent rates:

Event	Price	When it fires
apify-actor-start	$0.01	Once per MCP session (system event — fires when your client connects)
tool-call	$0.05	Every successful tool invocation
enrichment-tier	$0.10	When the tool returns enrichment data (emails / phones / profiles / funding) — fires for find_company_emails, enrich_company, aggregate_company_profile, find_b2b_leads, enrich_lead_list, track_startup_funding

Typical session math:

• Quick lookup (one find_company_emails call): 0.01 + 0.05 + 0.10 = $0.16
• Standard prospecting (5 tool calls, 3 enrichment): 0.01 + 5×0.05 + 3×0.10 = $0.56
• Heavy enrichment (10 calls, 8 enrichment): 0.01 + 10×0.05 + 8×0.10 = $1.31
• Bulk run (50 calls, 40 enrichment): 0.01 + 50×0.05 + 40×0.10 = $6.51

You only pay when a call succeeds and returns data — failed lookups don't charge the enrichment tier.

Why This Beats The Incumbents

Vendor	Entry price	Per-lookup cost	Bring-your-own-agent?	Coverage
Apollo.io	$99–$299/mo seat	$0.10–$0.30 per credit	Limited API	Contacts + companies
ZoomInfo	$14,995/yr seat	$0.30+ per record	Enterprise API only	Contacts + intent
Clearbit (HubSpot)	Enterprise quote	$0.50+ per enrichment	API yes, no MCP	Company + person
Lusha	$69+/mo seat	$0.10–$0.40 per credit	Limited API	Contacts only
Hunter.io	$49+/mo	$0.05 per find	API yes, no MCP	Emails only
Wappalyzer (RIP)	Was $149/mo	Subscription discontinued	n/a — service ended	Tech stack
Sales Intelligence MCP	$0 (PPE only)	$0.05–$0.15 per call	YES — MCP-native	All 10 use cases above

Three things make this different:

1. MCP-native — your AI agent calls the tools directly, not you clicking through a UI.
2. Per-call billing — no minimum, no seat, no 12-month contract. Run it once a week or 1,000 times a day.
3. Composable — chain the 10 tools in any order. Find → enrich → qualify → outreach, all inside one agent loop.

Don't Have an Apify Account?

Sign up free — you get platform credits on signup, and signing up via that link helps fund continued development of this fleet (we get a small affiliate cut, no extra cost to you).

Sister Actors

Use these alongside the MCP server, or hit them directly via the Apify API / SDK / Zapier / Make.com / n8n:

• Company Email Finder — bulk role-based + verified emails
• Company Enrichment Tool — per-company enrichment
• Company Data Aggregator — homepage + Crunchbase + JSON-LD aggregator
• B2B Lead Finder — multi-source B2B prospect discovery
• Lead List Enricher — bulk enrichment over domain lists
• Hiring Signal Detector — open-roles + growth-keyword scanner
• Startup Funding Tracker — Crunchbase News + TechCrunch funding feed
• Wappalyzer Replacement — drop-in tech-stack detector
• YC Directory Scraper — full YC company directory
• LinkedIn Jobs Scraper — public job-posting search

FAQ

Q: Do I need a separate Apify account for each upstream actor? No. One Apify token unlocks all 10 tools through the MCP server. The server orchestrates everything server-side using Actor.call(...).

Q: Can I use this in production agent workflows? Yes. Apify Standby mode keeps a warm container ready for incoming MCP sessions, with auto-scaling on the platform side. Tool calls are billed only on success, so retries don't double-charge.

Q: What's the typical latency? Tool calls range from 5s (cached enrichment) to ~120s (LinkedIn jobs with full descriptions, full YC batch dump). Most calls land in the 15-45s range. The MCP transport keeps the connection open while the upstream runs.

Q: Can I extend it with my own tools? The source is open in the actor — fork it on Apify, add tools, redeploy under your own namespace. The pattern is FastMCP @server.tool() decorators wrapping Actor.call(...).

Q: Is the upstream data quality good enough for outbound? Yes — these are the same actors used by ~400 paying NexGenData customers across 11 lead-gen actors. Email patterns are verified via DNS MX records; enrichment includes confidence scores where applicable.

---

Built and maintained by NexGenData — 165+ actors covering scraping, enrichment, MCP servers, and automation.

Home: thenextgennexus.com