About The Bombay Data Company

Who builds it

Kaavish Ramchandani, based in Mumbai. The company is Mask Management, a sole proprietorship of Kaavish Ramchandani, trading as The Bombay Data Company. The parent brand is Mask360, a creative and performance agency in Bombay. Grievance officer: Kaavish Ramchandani, connect@mask360.agency.

What it is

A 3D map of Bombay with 598 companies on it, brands, agencies, hotels, developers, retailers and funds, each with its nearest station and the walk in minutes, the commute and the rent nearby. It is built for the moment before someone applies anywhere: to see who is where and what the commute is.

Around the map: a page for every company, area, sector, station and rail line (browse them), a jobs page with every open role in the feed, and The Bombay Salary Index, which collects anonymous pay reports and publishes an area once 200 people have added theirs.

How the data is gathered

Company rows come from the company's own pages, from OpenStreetMap points with the same name and from a commercial enrichment vendor, and an address is never guessed: 366 from a commercial enrichment vendor, whose street geocode counts as exact only when it sits within 4.5 km of the row's own pincode and is otherwise placed by pincode or area, 124 from the company's own website, 63 from no source recorded on the row, 37 from an OpenStreetMap point with the same name and 8 from another primary source. 130 rows are marked source checked.

Every pin carries its precision. 294 rows sit on an exact street address, 113 on the centroid of the registered pincode, 104 on the area the company itself names, and 87 are listed without a fixed area because no source gave one.

The walk to the nearest station is a straight line estimate at 5 km/h from the pin, so the streets add time. The rail network the map draws is Western, Central, Harbour, Metro 1, Metro 3 and Metro 7.

Roles are read every hour from the employers' own hiring systems, Workday (229), Keka (75), Greenhouse (57), SmartRecruiters (41) and Lever (19) as of 23 September 2026, by a job that pulls each board and keeps titles, links, the posting date and the function. Nothing is copied from another job board.

What it refuses to do

It never guesses an address. An early audit found that most areas guessed from a company name were wrong, so a row with no source for its location is dropped or listed without a fixed area: 87 of the 598 today, never shown as located anywhere.

It carries no JobPosting markup, because it holds titles and links and not the full posting. It holds no CVs and has no candidate database, by design. Applying always goes to the employer.

It publishes no salary figure until 200 people have added theirs. A revenue figure on a company page is a third party estimate and is labelled as one. Every number on a generated page is counted from the dataset at build time and dated.

How to correct an entry

Write to connect@mask360.agency with the page and what is wrong. A wrong listing is corrected or removed within seven days, and the row is marked as checked. If you represent the company and want the listing changed or removed, the same address does it, with no form and no fee.

A company that is missing can be added for free; it goes up the same week.

How to cite

Cite the page you used, with the date the dataset carries. Every page says "as of" with its date in the first lines.

The Bombay Data Company (Mumbai, India), https://bombaydata.edgeone.cool/, dataset as of 23 September 2026.

For one page: The Bombay Data Company (Mumbai, India), "113 companies in Andheri East", https://bombaydata.edgeone.cool/in/andheri-east.html, dataset as of 23 September 2026. Machine readers can fetch the same facts from data/summary.json, llms.txt and llms-full.txt.

When things update

Roles: every hour, from the employers' own boards. The roles date on a page is the feed's own timestamp, named separately from the dataset date because it is usually a day or two behind.

Company rows: on each build of the site, and every page names its dataset date. The sitemap's lastmod moves only when the words on a page change, not on every build.

AI crawlers and machine readers

Crawling to index, answer and cite is welcome. robots.txt allows every crawler and names GPTBot, OAI-SearchBot, ClaudeBot, anthropic-ai, PerplexityBot, Google-Extended, Applebot-Extended and CCBot explicitly with the same allow. What we ask in return is the citation above, with the page and the date.

For machines: llms.txt says what the site is, llms-full.txt carries the answer block of every list page, data/summary.json is the fact sheet and jobs.xml is the roles feed. Bulk collection of the listings to republish them as a directory is not permitted under the terms; the dataset is licensed through Partners. Location data derived from OpenStreetMap stays under the ODbL.