Five patterns. One key.
Honest credit math on each.

The workflows builders actually ship on RichAPI — with the endpoints, the code, and what it costs.

01 Find active buyers every week (signal-based prospecting)

People who did something this week
beat any static list.

Pattern: pull post/hiring signals → enrich authors → find verified emails → hand to your outreach tool.

Credit math: ~80 credits/day ≈ $1.60 for 25 fresh, signal-qualified prospects.

ROI math: ≈ $0.09 per signal-qualified contact with a verified email — vs. 2–4 hours of manual LinkedIn + verifier work, or a $185+/mo platform seat.

Owns this internally: growth lead / RevOps.

MCP version: one prompt: "Find 25 people who posted about data pipeline migration this week, enrich, get verified emails, table the results."
cron.py — 6am
# 6am cron: who posted about {topic} in the last week?
posts = post("/post_keyword_search", {"keywords": "data pipeline migration",
        "datePosted": "past-week"})                       # 6 cr
authors = [post("/enrich_profile", {"url": p["author_url"]})
           for p in posts[:25]]                            # 25 cr
emails  = [post("/email_finder", {...}) for a in authors]  # ≤50 cr
# push the hits to your sequencer via its API — your code, your rules

02 Let your AI assistant do account research (agent-driven workflows)

The model decides the steps;
RichAPI is the action surface.

Pattern: account research, competitor mapping, event-attendee mining, post-engagement harvesting (post_activities on a viral post in your niche → enrich the commenters → emails).

example transcript frame
> prompt
richapi_post_activities      → 3 cr
richapi_enrich_profile ×20   → 20 cr
richapi_email_finder ×20     → 14 hits = 28 cr · 6 soft misses = 6 cr
→ ranked table with receipts

"57 credits · 14 verified emails · 6 misses billed at half."
ROI math:a research run that took an afternoon now costs ~$1.15 and runs while you're on a call.

Owns this internally: whoever runs outbound — founder, growth lead, or the one AI-curious AE.

Best for: AI-native founders, GTM builders prototyping agents. Sandbox covers the first experiment.

03 Build verified lists at agency scale (bulk jobs + webhooks)

$52 per thousand verified contacts.
No babysitting.

Pattern: people_search (0.1 cr/result — pay for what returns) → bulk email_finding jobs with webhook callback → email_verification before any send. Free inquiry endpoints poll job status at $0.

Credit math for a 1,000-lead book: search ~100cr + bulk email finding ~2,000cr (2cr per person on bulk jobs) + bulk verification ~500cr (0.5cr per email) ≈ 2,600 credits ≈ $52 per thousand verified contacts — agencies typically resell this at 10–20×. Bulk endpoints + webhooks mean no babysitting. Owns this internally: agency ops lead / delivery manager.

04 Add enrichment as a product feature (platform backend)

Your SaaS needs "enrich this user/company"
as a feature.

Pattern:you don't build a data team — you call person_enricher / company_enricher / web_tech_stack server-side, with our circuit breakers and provider failover absorbing vendor chaos.

ROI math: ship the feature in ~2 weeks of integration vs. ~6 engineer-weeks ($48K+) building the data layer in-house. Owns this internally: VP Product / platform engineering lead. Best for: sequencers, CRMs, vertical SaaS, AI-SDR platforms. → /partners

05 Score every lead against your ICP automatically (ai_enrich)

Your ICP definition lives in a prompt
your repo versions.

Pattern: ai_enrich runs GPT/Claude/Gemini server-side over each enriched record and returns a structured score. No second LLM bill, no glue code.

Credit math: enrich (1) + score (2) = 3 credits ≈ $0.06 per scored account.

ROI math: your reps stop manually triaging lists — $6 scores 100 accounts against criteria you control.

Owns this internally: RevOps.

MCP version: Claude applies the ICP inline — zero extra credits.
score.py
score = post("/ai_enrich", {
  "provider": "anthropic",
  "prompt": ICP_PROMPT,          # versioned in your repo
  "data": enriched_profile
})                                # 2 credits

What these have in common

One key. One credit pool. Receipts on every call.
And the miss never costs full price.

FAQ

Frequently asked questions

Pick a pattern. Ship it this afternoon.