Hiring-Intent Buying Signals

Turn public ATS job boards (Greenhouse, Lever) into B2B buying signals. When a company opens several roles in an area, it's investing there — a lead for whoever sells into that area.

Who buys this

B2B sales teams, SDRs, staffing/recruiting agencies, cold-email marketers, consultants. Hiring intent is a proven paid category (Apollo/Saleshandy $49–79/mo, IntenTrack $500/mo, Bombora/6sense $15–50k/yr).

What it does

1. Reads each company's public board (Greenhouse content=true, or Lever) — no auth.
2. Keeps roles matching your roleKeywords (word-boundary match, so "ai"/"ml" don't hit "maintenance").
3. Counts recent matching roles; if ≥ minOpenings, emits a buying signal with the top department.

Works out of the box

No need to know ATS board slugs: pick a companyPreset (ai, fintech, devtools, saas, marketplace, health, hr, all) — 674 company boards, each verified live (Greenhouse + Lever + Ashby, auto-routed). Just hit Start. Example: preset ai → 1,391 jobs (Anthropic 411, Scale AI 191, Databricks 789).

Three modes

• jobs — one row per job listing. The classic scraper.
• signals — per-company hiring intent: N open roles in an area = active investment = a lead.
• techstack — technographics: which tools a company actually uses, extracted from its own job descriptions with a deterministic dictionary (no LLM, no guessing — every hit is verifiable in the post). Example: Databricks → Spark, AWS, Azure, Python, Scala. Use it to find companies running a stack you sell into.

What the technographics data actually claims

"Technologies mentioned repeatedly in this company's own job posts" — not "technologies certainly running in production". Every hit is verifiable: it sits in a public job ad you can open. Single mentions are filtered out by default (minTechMentions: 2), because one mention is usually a nice-to-have or a migration target, not the stack.

Validated on companies with publicly known stacks: 14/14 expected technologies detected (Databricks→Spark/Scala, GitLab→Ruby/Kubernetes, Vercel→Next.js, Cloudflare→Rust/Go). No implausible claims on the hard cases (MongoDB shows no competing database; Datadog shows no Grafana).

The money query: competitive displacement

techstack mode + usesTech / notUsesTech builds the list a competing vendor pays for:

uses Datadog, does not use Grafana → 18 companies (Coinbase, Chime, Block, Marqeta, Tide, Vercel…)

Every tool detected is also a vendor whose competitors want that list. Same for gaps: categoryGaps shows categories where a company shows no tool at all — an open slot to sell into. Gaps are only reported when enough job text was analysed (gapsReliable), so absence means absence, not "we didn't look".

Input

Field	Meaning	Default
companyPreset	curated verified boards (use if you don't know slugs)	all
companies	your own ATS board tokens; overrides preset	—
locationFilter	keep jobs whose location contains this (e.g. Berlin, Remote)	—
remoteOnly	keep only remote-flagged jobs	false
departmentFilter	keep jobs whose department contains this (e.g. Engineering)	—
maxJobsPerCompany	cap results per company	0 (no cap)
outputMode	jobs (one row per listing), signals (per-company hiring intent), techstack (tools a company uses)	jobs
source	greenhouse or lever	greenhouse
roleKeywords	roles that signal the investment you sell into	sales/AE/SDR/revenue
minOpenings	surge threshold (≥ N matching roles)	2
sinceDays	only roles updated within N days	30

Output (dataset)

company, source, openingsMatched, topDepartment, roles[], signalReason, boardUrl.

Example

roleKeywords: ["machine learning","ml","ai"] → finds companies scaling AI teams → sell them AI infra / data tools / ML consulting. (Tested live: Stripe 40, Figma 21, GitLab 18 AI/ML roles.)

Notes

• Public endpoints only, rate-limited politely. No personal data (public job posts).
• Pay-per-event: charges per signal (signal event) once monetization is configured in Console.

Getting board tokens

Greenhouse: the slug in boards.greenhouse.io/<token>. Lever: jobs.lever.co/<token>. Thousands of companies use these — build target lists per industry.