Every workspace opens with datasets built for investment questions, kept as the raw material other providers summarize away, and the connections to bring in what your firm already has.
Which models and labs developers are paying for, day by day
Where an app or product ranks in the App Store, Google Play, and Amazon
How often the internet looks up a domain, ranked daily
Start research from your inbox and get digests and alerts back
Hundreds of thousands of economic series from the St. Louis Fed
What developers build with, read from public commits and pull requests
What an hour of H100, H200, or B200 compute costs, across markets
Earnings calls, slide decks, and releases, as text
Your workspace’s data and monitors, from the assistants you already use
Shows and episodes across the open podcast ecosystem, searchable
Odds, volume, and open interest from Kalshi and Polymarket
Daily downloads for any Python package, split by operating system and CI
Trips, drivers, and robotaxi miles from the regulators that collect them
A site’s progress from orbit, as a timeline you can revisit
Every SEC filing, in full text
Stablecoin supply by chain, and how much of it is actually moving
The vendors software companies list on their trust centers
The technology behind 25 million websites, kept as raw evidence
How any page looked on any date, back to the 1990s
Searchable posts from founders and analysts on X
Connect the APIs your firm licenses, or ones you operate, with a key
Nothing matches. Clear a filter or try another word.
The technology behind 25 million websites, kept as raw evidence.
A weekly crawl of 25.7 million hosts: every origin in Google’s Chrome UX Report plus five million responsive hosts from Cloudflare’s rankings. Village keeps what the browser saw rather than a vendor label, so detection is written as SQL at query time and you decide what counts as evidence.
Shows and episodes across the open podcast ecosystem, searchable.
A catalog of shows and episodes from public RSS feeds, built for the questions that only make sense once a language model can read show notes and transcripts: who appeared where, and what they said before the market cared.
The vendors software companies list on their trust centers.
Subprocessor disclosures from public trust centers, read by the agent when you ask, so you can see which vendors a company relies on and count them across the companies you follow.
How any page looked on any date, back to the 1990s.
The Internet Archive’s Wayback Machine, reachable from the agent, so a pricing page, a careers page, or a customer list can be read as it stood on a given date and compared with today.
A site’s progress from orbit, as a timeline you can revisit.
Optical and radar scenes from public satellite catalogs, found and compared by the agent so visible change at a plant, a mine, or a data center sits beside the filings and announcements about it.
Every SEC filing, in full text.
The SEC’s filing archive, ingested as full documents rather than extracted fields. Agents list filings by company, form type, and date, then read the whole text or the rendered PDF.
Earnings calls, slide decks, and releases, as text.
Transcripts, presentations, and releases from public companies’ investor events, licensed from Quartr and processed so an agent can quote them. Slides are read with OCR; transcripts carry timestamps.
Hundreds of thousands of economic series from the St. Louis Fed.
The Federal Reserve Bank of St. Louis’s FRED database, read through its official API: employment, prices, housing, industrial production, and rates from more than a hundred government and private sources, so a company’s own signals can be set against the economy around it.
Which models and labs developers are paying for, day by day.
Token and spend shares from the public leaderboards of OpenRouter and Vercel’s AI Gateway, alongside Ramp’s AI Index of business spending, so a lab’s share can be followed both in API traffic and in the companies paying for it.
Daily downloads for any Python package, split by operating system and CI.
PyPI download counts from ClickPy, ClickHouse’s public copy of the PyPI download logs, queried per package and per day. Developers install an SDK before their usage shows up in revenue, so downloads make a leading indicator for the companies behind them.
What developers build with, read from public commits and pull requests.
Public GitHub activity from GH Archive’s hourly event record and GitHub’s search API, so adoption of a framework, a language, or a developer tool can be counted in the repositories where it shows up.
How often the internet looks up a domain, ranked daily.
Cisco Umbrella’s daily ranking of the million most-queried domains, kept as dated archives so a product’s domains can be followed from launch. It reaches apps and APIs that have no website traffic to measure.
Where an app or product ranks in the App Store, Google Play, and Amazon.
Chart positions, ratings, and review counts collected from public store pages and chart feeds, so a launch’s traction or a product’s sell-through can be followed daily instead of waiting for the company to report it.
What an hour of H100, H200, or B200 compute costs, across markets.
Rental prices for data-center and consumer GPUs from public marketplaces and indices: Ornn’s index of executed rentals, Vast.ai’s live offers, and Kalshi’s contracts on future prices, so the cost of compute can be followed chip by chip as new generations arrive.
Stablecoin supply by chain, and how much of it is actually moving.
Circulating supply from DefiLlama, issuer attestations from Circle, and adjusted transaction volume from Visa Onchain Analytics, so growth in digital dollars can be split into supply that sits and supply that moves.
Odds, volume, and open interest from Kalshi and Polymarket.
Contract prices and trading activity from Kalshi’s and Polymarket’s public market-data APIs, so the market’s probability on an election, a rate decision, or a product launch can be tracked as a series, and the venues’ own growth measured.
Trips, drivers, and robotaxi miles from the regulators that collect them.
The trip and fleet records that cities and states require ride-hail and autonomous-vehicle operators to file, from New York’s Taxi and Limousine Commission, Chicago’s open data portal, and the California Public Utilities Commission, so Uber, Lyft, and Waymo volumes can be read from filings rather than estimated.
Searchable posts from founders and analysts on X.
Posts and accounts from X, brought into the workspace so an agent can search what was said, track who said it, and monitor the accounts that matter to a thesis.
Your workspace’s data and monitors, from the assistants you already use.
Village’s MCP server puts a workspace’s files, datasets, and monitors to work inside Claude and ChatGPT, so research continues in whichever conversation you are already in.
Start research from your inbox and get digests and alerts back.
Every workspace has an address. Email it a question and the agent replies in the thread; monitors send daily digests and alerts the same way.
Connect the APIs your firm licenses, or ones you operate, with a key.
Add API keys for the data services you already pay for, and agents use them in research and recurring monitors without asking for the key again.
