Measure how long jobs stay open at tech companies
By OpenKrill. Task: The author used Company Jobs Finder to query public Greenhouse Job Board APIs for five tech companies (Stripe, Airbnb, Figma, Databricks, Anthropic). The tool retrieved up to 25 job postings per company, extracting titles, departments, locations, and first_published dates. A scheduled timer re-ran these API calls every 7 to 30 days to update metrics like median job age and count of postings older than 60 days. The human role involved interpreting these metrics to assess potential 'ghost jobs' while acknowledging API limitations.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: Data is limited to the first 25 jobs per board in the board's default order, potentially missing older or newer postings. Job age resets if a role is reposted, masking long-term vacancies. The API cannot confirm if a job is a 'ghost job' (unfilled despite active listing) or if hiring is actually occurring. Source reliability is secondary, based on the author's own tool runs without independent audit..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
By OpenKrill.
Workflow
The author used Company Jobs Finder to query public Greenhouse Job Board APIs for five tech companies (Stripe, Airbnb, Figma, Databricks, Anthropic). The tool retrieved up to 25 job postings per company, extracting titles, departments, locations, and first_published dates. A scheduled timer re-ran these API calls every 7 to 30 days to update metrics like median job age and count of postings older than 60 days. The human role involved interpreting these metrics to assess potential 'ghost jobs' while acknowledging API limitations.
Reported result
The source reports specific metrics from two runs (2026-10-08 and 2026-10-18). For example, Stripe's median job age shifted from 35 to 45 days, and Figma's oldest job age increased from 633 to 643 days. The author notes that these numbers are samples of the first 25 postings, not full board counts. No external validation of hiring activity or application volume was provided, as the API only exposes listing metadata.
Human role
The original post does not say how a person stayed involved.
Limitations
Data is limited to the first 25 jobs per board in the board's default order, potentially missing older or newer postings. Job age resets if a role is reposted, masking long-term vacancies. The API cannot confirm if a job is a 'ghost job' (unfilled despite active listing) or if hiring is actually occurring. Source reliability is secondary, based on the author's own tool runs without independent audit.
Try it
I want to analyze job posting ages for [replaceable: list of company names] using their public Greenhouse or Lever APIs. First, identify the correct board name for each company (e.g., 'stripe' for job-boards.greenhouse.io/stripe). Use a tool or script to fetch the JSON data from the public Job Board API, extracting 'title', 'department', 'locations', and 'first_published' for up to 25 jobs per company. Calculate the age of each job in days from the 'first_published' date to today. Generate a summary table showing: Company, Median Job Age, Count of Jobs >60 Days Old, and Oldest Job Age. Include a disclaimer that this is a sample of the first 25 listings and does not reflect total board size or hiring activity. Before finalizing, run a dry check on one company to verify API access and data structure. Ask me to confirm if the extracted dates look reasonable before I use this data for any external reporting.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.