license: cc-by-4.0
language:
- en
pretty_name: 'Jobspiq open data: India''s tech job market'
size_categories:
- n<1K
tags:
- india
- jobs
- job-market
- tech-jobs
- salaries
- skills
- labour-market
configs:
- config_name: skills_demand
default: true
data_files:
- split: train
path: data/*/skills-demand-india.csv
- config_name: tech_salaries_stated
data_files:
- split: train
path: data/*/tech-salaries-stated-india.csv
Jobspiq open data: India's tech job market
Free, citable data from Jobspiq, an AI job search platform for technology professionals in India.
Each folder in data/ is a dated snapshot of the datasets published at jobspiq.in/data. The website shows live figures that change every day; the files here never change once committed, so you can cite an exact version. The same snapshots are on GitHub at jobspiq/jobspiq-public.
Datasets
| Config | File | What it contains | Window |
|---|---|---|---|
skills_demand |
skills-demand-india.csv |
Active tech job postings in India that mention each skill, and each skill's share of all tech postings | Postings seen in the 30 days before the snapshot date |
tech_salaries_stated |
tech-salaries-stated-india.csv |
Quartiles of salaries that employers state in tech job postings (no estimates), by experience, role family and skill, with the number of stated salaries behind every row | Postings seen in the 90 days before the snapshot date |
Each config combines every snapshot; filter on the as_of column to pick one date. Column definitions are in data/README.md.
from datasets import load_dataset
skills = load_dataset("jaydeepgio/Jobspiq", "skills_demand", split="train")
salaries = load_dataset("jaydeepgio/Jobspiq", "tech_salaries_stated", split="train")
The data covers active technology job postings located in India, or remote and open to candidates in India. It contains aggregate numbers only: no job descriptions, no copied postings and no personal data.
Method and limits
- Postings are collected from LinkedIn, Naukri, remote job boards and employer career pages, and duplicates are merged. A job counts once per skill.
- Salaries are only those stated in the posting. Most postings state none, so every salary row shows its sample size, and groups with fewer than 25 stated salaries are left out. Employers who publish pay are not a random sample of all employers.
- These are job postings, not hires, and they do not measure total employment.
Longer method notes for Jobspiq's published research are in methods/.
How to cite
Source: Jobspiq, "[dataset name]", snapshot [YYYY-MM-DD], https://huggingface.co/datasets/jaydeepgio/Jobspiq. Licensed CC BY 4.0.
@misc{jobspiq_open_data,
author = {{Jobspiq}},
title = {Jobspiq open data: India's tech job market},
howpublished = {\url{https://huggingface.co/datasets/jaydeepgio/Jobspiq}},
note = {Also at https://github.com/jobspiq/jobspiq-public. CC BY 4.0}
}
Licence
The data is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0): you may share and adapt it for any purpose, including commercially, as long as you credit Jobspiq.
Contact
Questions, corrections or press enquiries: support@jobspiq.in · Press kit: jobspiq.in/press