State agencies publish their registries in different forms: Florida as fixed-width text files; Colorado, New York and Texas through open-data portals. Our pipeline downloads each source on a fixed morning schedule, parses and checks the records, skips records it already holds, and publishes each dataset as a CSV file at a fixed URL, rebuilt each morning from the source's latest release (the Florida corporate datasets Monday to Saturday). You subscribe to the files; we run the pipeline.
Try any feed free for 7 days, no card. We email you one link to the full current file(s) of the feed you pick, the same files a subscriber gets; it works for 7 days. During the trial we send three short emails (day 1, day 5 and day 7), each with a one-click unsubscribe; then nothing. One free trial per email address and per network. We keep your address, the feed you chose and your network prefix until 90 days after the trial ends, then delete them. Privacy.
Sent on Mondays: last week’s new-filing counts and a link to the current free sample (the water edition: the changes we detected and links to the three free radar samples). Confirm by email; unsubscribe in one click.
The records are public. Making them usable is the work — and it is not trivial.
Florida publishes its corporate registry as fixed-width ASCII: 1,440-character filing records with packed officer blocks, and 662-character event records. We parse them from the state's published layouts, corrected where live records differ, check every record's length and key fields, and quarantine a whole file whose records do not match.
Entity types and status codes are kept exactly as each state codes them, so they differ between states. Where a Florida record has no state, it is filled in from the ZIP code. A Florida file that fails its checks is set aside and reported to us, so a bad source day shows up as a delay rather than as a half-loaded file.
States sometimes re-post the same file under a new date: Florida's corporate event file repeated the previous day's contents on 10 of the 59 days to 10 October 2026, about one day in six. We skip records we already hold, matched on the state's own identifier, so a re-posted file does not show up as new rows.
Florida publishes a daily file of corporate events coded in its own vocabulary. We map those codes to plain event types — dissolutions, reinstatements, amendments, conversions, name changes, registered-agent changes — and attach each entity's name from our copy of the state's full registry. For the water datasets we compare each new copy of the EPA and Florida DEP data with the previous one and log what changed.
Florida's corporate filing and event files are archived with a SHA-256 checksum before parsing, so any Florida registry row can be traced to the state file it came from. For the other sources we keep the latest copy we downloaded, without a checksum.
Every file has a header row, and column names rarely change: the one change since launch removed two tax-ID columns from the fictitious-names file on 15 September 2026. Most files are written under a temporary name and then swapped in whole. The Florida Business Filings and Florida Registry Change Feed files are copied into place instead, so a download made during that copy could be incomplete.
CSV files at fixed URLs. Each is rebuilt each morning from the source's latest release, on the days shown on its card. Month-to-month, cancel any time.
A different product line, and a different kind of work. Three EPA datasets joined on PWSID — the public water system identifier — plus Florida DEP State Revolving Fund (SRF) projects matched to systems by sponsor name. Each system gets a trigger score: a 0–100 prioritisation index built from published records, not a prediction. The water datasets describe water systems, not people.
Try either water feed free for 7 days, no card: the full files, by email.
Browse Florida’s community and non-transient non-community water systems — reported lead and galvanized lines, PFAS results and federal deadlines, one page per system.
Florida Business Filings, one entity per row (invented example values):
| field | example |
|---|---|
| doc_number | L26000000000 |
| entity_name | EXAMPLE ENTITY NAME, LLC |
| entity_type / status | FLAL · A |
| filing_date | 2026-10-02 |
| principal address | 100 EXAMPLE ST, TAMPA, FL 33602 |
| registered agent | EXAMPLE REGISTERED AGENT, INC. |
| first_seen | 2026-10-09 |
Subscriptions are month to month, cancel any time. Files are delivered as downloads at fixed URLs.
These rules apply to every FilingsDaily dataset. Use them for research, compliance, and business-to-business outreach — within the law.
Official data published by state agencies — the Florida Department of State, Division of Corporations (its bulk-data SFTP server, which uses a public login the state publishes, and its public fictitious-name file site); and the Colorado Department of State, the New York Department of State and the Texas Comptroller of Public Accounts, through their open-data portals. The water datasets come from EPA and the Florida Department of Environmental Protection (see below). We do not scrape web pages, and we do not access any non-public system. FilingsDaily is not affiliated with, endorsed by, or connected to any government agency.
Engineering, not access. Florida's files are fixed-width exports with coded fields and repeated postings, and each state portal has its own formats and vocabularies. Building and running a pipeline that parses them, skips records it already holds, turns Florida's event codes into plain events, scores the water systems, and rebuilds each file each morning from the source's latest release is the product. If you would rather build that yourself, the sources are public and we will happily point you at them.
Every dataset is rebuilt each morning from the source's latest release. Each new-entity file holds the source's newest release only: if a source posts a day late, or later adds rows to a day it has already released, those rows are not added to your file. Change files hold a rolling window: 7 days for the Florida Registry Change Feed, 90 days for each water change log. Florida Business Filings and the Florida Registry Change Feed: rebuilt Monday to Saturday; Florida posts its files on weekday evenings, so new rows arrive Tuesday to Saturday, and Monday brings no new rows. The change file holds the last 7 days of events because the state's event file does not change every day. Florida Fictitious Names and Colorado Business Filings: rebuilt every morning. New York Business Filings: rebuilt every morning; the newest formation date is usually two days old. Texas is different and we say so on the product: it comes from the state franchise tax roll, where an entity appears roughly three weeks after it is chartered, so the newest charter date in the Texas file is about 21 days old, by design rather than by fault. Water datasets: rebuilt every morning from our latest copies of the EPA and Florida DEP data; we re-check the SRF projects daily, the EPA Service Line Inventory and the AWIA certifications weekly, and UCMR 5 monthly.
Plain CSV with a header row, UTF-8, one row per entity, owner, water system or event depending on the dataset. Column names rarely change; the one change since launch removed two tax-ID columns from the fictitious-names file on 15 September 2026. Most files are swapped in whole; the Florida Business Filings and Florida Registry Change Feed files are copied into place, so a download made during that copy could be incomplete.
No telephone numbers or email addresses: none of our sources contain them. Names and addresses, yes, as filed with the state: registered agents (names in Florida; names and addresses in Colorado and New York), fictitious-name owners (Florida), service-of-process contacts (New York) and the entities’ own addresses. See Privacy.
Pick a feed and give us your email address, and we email you one link to the full current file(s) of that feed, the same files a subscriber gets. The link works for 7 days. There is no card and nothing to cancel: the trial simply ends. During the trial we send three short emails, on day 1, day 5 and day 7, each with a one-click unsubscribe, and nothing after it. One free trial per email address and per network. The Terms of service apply during the trial. To keep the data after the trial, subscribe from the feed’s card.
Subscriptions are month to month and can be cancelled at any time from the link on your Paddle receipt or at paddle.net. You can also email support@filingsdaily.com and we will cancel it in Paddle for you.
Paddle.com Market Limited is the merchant of record and payment processor for all purchases on this site, and refunds are handled under Paddle's refund policy. You can request a refund directly from the receipt Paddle emails you after purchase, or via paddle.net. We're also glad to help — email support@filingsdaily.com and we'll point you to the right place or flag anything unusual to Paddle on your behalf.
Three EPA datasets, joined on PWSID: the EPA Service Line Inventory (SDWIS Federal Reports); EPA UCMR 5 PFAS occurrence results; and EPA's AWIA certification data. Florida DEP's Drinking Water SRF project data is matched to systems by sponsor name, and the federal compliance deadlines come from the rules themselves. Each source is retrieved from its official public endpoint. The trigger score, the sponsor-name match and the change logs are ours.
The trigger score is a 0–100 prioritisation index built from published records, not a prediction. In the Florida Lead Service Line Radar it adds four parts: regulatory pressure (up to 40: line counts up to 34, plus 6 for a lead action-level exceedance, which EPA's file leaves blank for every Florida system), SRF project match (up to 25: a project record of any status +10, an active project +5, principal forgiveness +5, $1M or more +5), deadline proximity (up to 25: 15 for every system on 10 October 2026, rising as the 2027-11-01 deadline approaches) and recent change (up to 10; 0 for every system as of 10 October 2026). The PFAS radar adds detection, population, SRF funding and deadline parts; the resilience radar adds certification-gap, deadline, system-size and SRF funding parts. Every part is a column in the file, so you can check the reasoning or re-weight it yourself.
Sometimes. Each new state needs its own parser, so tell us what you need and we will tell you whether we can do it.
What we collect from you, and separately, what the datasets contain. Those are two different things.
Your name and business email, so we can identify your subscription and send access and renewal notices. Billing details, which go directly to our payment processor — we never see or store card numbers. Correspondence you send us. Paid-file delivery logs — the network prefix of the requesting IP address (never the full address), the time, and the file requested — for 90 days, to operate the service, diagnose failed downloads, and detect shared download links.
Free samples by email. If you ask for a free sample, we keep your email address, the sample you chose, when you asked, and the network prefix of your IP address (never the full address) for 90 days, to send you the link and to stop abuse, and then delete them. The email carries one link, which works for 24 hours; we record when it is used. We send you nothing else unless you also tick the box for the weekly email. If you tick the weekly-email box, the same email carries that list’s confirmation link, and nothing weekly is sent until you click it.
The free trial. If you start a free trial, we keep your email address, the feed you chose, when you asked, the version of the terms you accepted, and the network prefix of your IP address (never the full address), until 90 days after the trial ends, and then delete them. We record when your link is used. To allow one trial per address and per network, we keep a one-way hash of your email address for one year. We send you the link and three emails during the trial (day 1, day 5 and day 7), each with a one-click unsubscribe, and nothing after the trial; a short notice that a trial started (the feed and your address’s domain, not the address) goes to our own operations chat.
The free weekly email. If you sign up, we keep your email address, the list you chose, when you signed up and confirmed, and a hash of the IP address you signed up from. We send nothing except the confirmation email until you click its link, and every weekly email carries a one-click unsubscribe. After you unsubscribe we keep your address, marked as unsubscribed, so that we never email it again; email us if you want it deleted entirely. Free sample downloads are counted by file, network prefix, whether the link came from the sample form or the weekly email, and, if your browser sends one, the referring site. Our public pages server (the water and new-business pages, the sign-up form and the free samples) also writes each request, including the full IP address, to its system log, which is kept until the log reaches its size limit.
We have no customer account system, so we hold no passwords. We do not sell, rent, or share customer information, we run no advertising or cross-site tracking, and we do not send marketing email to people who have not asked for it.
The datasets are derived from records that state and federal governments publish as public records. They are not compiled from consumer data, purchased lists, or tracking.
Water infrastructure datasets are keyed on PWSID and describe water systems, not people. A few system names, as EPA publishes them, include an individual’s name (for example, a farm labour camp named for its owner); we add no personal data.
Business registry datasets describe registered entities. Because state corporate and fictitious-name filings are public by law and name the people a business lists in them, these datasets do contain personal data. Specifically: names of registered agents, including where the agent is a natural person rather than a company, and the agent’s address in the Colorado and New York datasets; names and addresses of New York service-of-process contacts; names and street addresses of business owners, where a fictitious-name (DBA) filing lists a natural person as owner; and the principal and mailing addresses of the entity itself, which for sole proprietors and small companies are frequently residential addresses. The Texas dataset also carries each entity’s Texas taxpayer number as the Comptroller publishes it. Apart from our own processing columns (such as first_seen and extracted_at) and a blank Florida state filled in from the ZIP code, these are the fields as the state publishes them. Our sources contain no telephone numbers and no email addresses, and we add none.
We do not enrich, append, cross-reference against consumer databases, infer, or score individuals. We republish the records as the state publishes them and derive business events — never personal attributes.
The underlying record is a public filing held by the issuing agency, and we publish what that agency publishes. We cannot amend a government record — corrections are made at the source. Our files show each record as the source published it on the day we took it, so a correction appears in a later file only if the source publishes it as a new filing or event. If you believe we published something the source does not say, email us and we will check it against our copy of the source file (for Florida corporate files, an archived copy with a SHA-256 checksum).
We use Cloudflare for DNS, delivery and cookieless page-view counts (Cloudflare Web Analytics), Paddle as merchant of record for billing, and an email provider for notices and support. They process data on our instructions; we do not sell data to or through them.
Customer records are kept for the life of the subscription plus seven years for tax purposes. Paid-file delivery logs are kept 90 days. Archived Florida corporate source files are retained indefinitely with SHA-256 checksums, because dataset auditability depends on it; for other sources we keep the latest copy we downloaded.
To access, correct, or delete your customer data, email support@filingsdaily.com and we will respond within 30 days. Records we are required to keep for tax purposes stay for the seven years stated above even after a deletion request. We delete the rest of your customer data from our working systems; copies that remain in server logs and backups are removed only as those logs rotate and those backups are replaced. We operate from the United States; our service providers (Paddle, Cloudflare and our email provider) may process data in other countries.
Business-to-business terms, version 2026-10-20. You agree to them by ticking the box before checkout, or on the free-trial form.
A subscription to one or more datasets, delivered as CSV files at fixed URLs and rebuilt each morning from the source's latest release on the schedule described for each dataset. Each new-entity file holds the source's newest release only; change files hold a rolling window (7 days for the Florida Registry Change Feed, 90 days for each water change log). You are paying for the derivation work: parsing source files, checking records, skipping re-posted records, turning Florida's event codes into change events, comparing successive copies of the water data, scoring water systems, and keeping column names stable. The underlying government records are public and free; the datasets are not the records. We are not affiliated with, endorsed by, or acting on behalf of any government agency.
Free trial. You may try any dataset free for 7 days by email, without a card: we email a link to the dataset’s current files that works for 7 days and then stops. One free trial per email address and per network; we may refuse or end a trial that breaks that rule or these terms. The licence and every restriction in these terms apply during the trial exactly as they do to a subscription, and you accept them by ticking the box on the trial form.
Subscriptions run month to month and renew until cancelled. Cancel any time from the link on your Paddle receipt or at paddle.net, or by email. Prices are as listed, in US dollars; depending on your location, Paddle adds tax to them or includes it in them, and shows the total before you pay; we may change prices for future renewals with at least 30 days’ notice, and a price you have already paid never changes. Paddle is the merchant of record and handles all billing and refunds under Paddle’s refund policy — see refund terms below.
Government sources publish on their own schedules. Each new-entity file holds the newest release only, so a day on which the source does not publish, posts late, or later adds rows to a day it has already released is not a failure of delivery. Persistent non-delivery is.
While your subscription is current you have a non-exclusive, non-transferable licence to use the datasets internally for business research, market analysis, compliance, due diligence, and record-keeping, including in derived analysis you produce for your own business or clients.
No FCRA use. The datasets are not consumer reports and we are not a consumer reporting agency. You must not use them for any purpose covered by the Fair Credit Reporting Act, including decisions about credit, insurance, employment, housing, or tenant screening, or any comparable purpose under state law.
Lawful contact only. You may use the datasets to contact the businesses they list. If you do — by mail, email, telephone, text, or otherwise — you are solely responsible for complying with the CAN-SPAM Act, the Telephone Consumer Protection Act, the National Do Not Call Registry rules, and applicable state law, including honoring every opt-out. Our sources contain no phone numbers or email addresses; appending contact data from elsewhere and contacting anyone is your act and your liability.
No bulk redistribution. You may not republish, resell, sublicense, or redistribute the datasets in bulk or as a substantially similar product. Producing your own analysis is permitted; reselling our files is not.
No unlawful use, and no credential or URL sharing to allow access by parties without their own subscription. Breach of this section terminates the licence immediately.
The datasets are provided as-is. We derive them from government sources that are themselves incomplete, inconsistently formatted, and occasionally wrong. We check records, quarantine Florida files that fail their checks rather than loading them half-read, and keep copies of the source files (with SHA-256 checksums for Florida corporate files) — but we do not warrant that the data is accurate, complete, current, or fit for a particular purpose, nor that availability is uninterrupted. Do not use these datasets as the sole basis for a decision with legal or financial consequences; verify against the issuing agency.
To the maximum extent permitted by law we are not liable for indirect, incidental, special, consequential, or punitive damages, or for lost profits, revenue, data, or business opportunity. Our total liability for any claim is limited to the fees you paid in the 12 months before the claim arose. You will indemnify us against claims arising from your use in breach of the section above.
These terms are governed by the laws of the State of Florida, with exclusive venue in the state or federal courts located in Florida. Material changes take effect for existing subscribers at the next renewal, with at least 30 days’ notice by email.