Job market analytics

1. 01Fix the roles, cities, and window

   Pick the search terms and locations once and never change them mid-series, or the counts stop being comparable. posted=7d keeps each weekly run to the openings posted since the last one.

   GET /v1/search?engine=indeed&q=software+engineer&location=New+York%2C+NY · query

   {
     "q": "software engineer",
     "location": "New York, NY",
     "job_type": null,
     "remote": false,
     "posted": null,
     "page": 1
   }

2. 02Read each search

   Each result carries the posted salary as min, max, and period, the job type, whether it is remote, and the posting date. Keep only yearly salaries for pay statistics; hourly postings answer a different question.

   GET /v1/search?engine=indeed&q=software+engineer&location=New+York%2C+NY · results.0

   {
     "position": 1,
     "id": "e6489bd83820f079",
     "title": "Software Engineer, Early Career, Campus",
     "company": "Google",
     "location": "New York, NY",
     "workplace_type": null,
     "remote": false,
     "salary_text": "$123,000 - $174,000 a year",
     "salary": {
       "min": 123000,
       "max": 174000,
       "median": null,
       "currency": "USD",
       "period": "year",
       "source": "employer"
     },
     "job_type": "Full-time",
     "posted_age": "30+ days ago",
     "posted_at": "2026-08-07T05:00:00.000Z",
     "snippet": "We're looking for engineers who bring fresh ideas from all areas, including information retrieval, distributed computing, large-scale system design, networking",
     "benefits": [
       "Paid holidays",
       "Dental insurance",
       "Health insurance",
       "401(k) matching",
       "Paid time off"
     ],
     "company_rating": 4.3,
     "company_reviews_count": 6269,
     "company_url": "https://www.indeed.com/cmp/Google",
     "easy_apply": false,
     "urgently_hiring": false,
     "sponsored": false,
     "url": "https://www.indeed.com/viewjob?jk=e6489bd83820f079"
   }

3. 03Open postings when you need the detail

   The job endpoint returns the full description, benefits, and when the posting expires. You do not need it for counts, but it is where skills and requirements are.

   GET /v1/job?engine=indeed&id=e6489bd83820f079 · job

   {
     "id": "e6489bd83820f079",
     "title": "Software Engineer, Early Career, Campus",
     "company": "Google",
     "location": "New York, NY",
     "workplace_type": null,
     "remote": false,
     "salary_text": "$123,000 - $174,000 a year",
     "salary": {
       "min": 123000,
       "max": 174000,
       "median": null,
       "currency": "USD",
       "period": "year",
       "source": null
     },
     "job_type": "Full-time",
     "posted_at": "2026-08-07T14:17:37.961Z",
     "valid_through": "2027-01-23T14:14:21.706Z",
     "company_logo": "https://d2q79iu7y748jz.cloudfront.net/s/_squarelogo/256x256/d778456838f54580d43f7dd78b3b0d9c",
     "company_url": "https://www.indeed.com/cmp/Google",
     "benefits": [
       "Parental leave",
       "Health insurance",
       "401(k) matching",
       "Paid time off"
     ],
     "direct_apply": false,
     "url": "https://www.indeed.com/viewjob?jk=e6489bd83820f079",
     "highlights": {
       "qualifications": [
         "Bachelor's degree in Computer Science, a similar technical field of study, or equivalent practical experience."
       ],
       "responsibilities": [
         "Design, develop, test, deploy, maintain, and improve software."
       ],
       "benefits": []
     },
     "paragraphs": [
       "Google is hiring early career software engineers in New York."
     ]
   }

# Weekly count of new openings and posted pay for a set of roles and cities.
# Appends to market.csv; chart it to see demand and pay move.
import csv, datetime, os, statistics, requests

API = "https://api.clair.im"
HEADERS = {"Authorization": f"Bearer {os.environ['CLAIR_API_KEY']}"}
ROLES = ["data engineer", "machine learning engineer"]
CITIES = ["New York, NY", "Austin, TX", "Seattle, WA"]

def search(q, location):
    r = requests.get(f"{API}/v1/search", headers=HEADERS, timeout=90, params={
        "engine": "indeed", "q": q, "location": location, "posted": "7d"})
    r.raise_for_status()
    return r.json()["results"]

week = datetime.date.today().isocalendar()
with open("market.csv", "a", newline="") as f:
    out = csv.writer(f)
    for role in ROLES:
        for city in CITIES:
            jobs = search(role, city)
            # Annual pay only; hourly and monthly postings are kept out of the median.
            mids = [(j["salary"]["min"] + j["salary"]["max"]) / 2 for j in jobs
                    if j["salary"] and j["salary"]["period"] == "year"
                    and j["salary"]["min"] and j["salary"]["max"]]
            remote = sum(1 for j in jobs if j["remote"])
            out.writerow([f"{week.year}-W{week.week:02d}", role, city, len(jobs),
                          len(mids), statistics.median(mids) if mids else None, remote])

Set CLAIR_API_KEY to a key subscribed to the Indeed Jobs API. Each call is one request against that API's monthly quota.

• One search returns the one page Indeed shows signed-out visitors, about 45 cards. Busy roles in large cities have more new openings than that per week; split them with job_type, remote, or narrower titles, or count daily.
• Indeed.com only, so United States postings.
• Salary appears only when the posting or Indeed shows one. The share with pay varies by state, since some states require pay ranges in postings.
• Postings are duplicated across boards and re-posted. Deduplicate on id within Indeed, and on company plus title across sources.