Products

WhenFree solves the specific scheduling problem that shared-calendar tools don’t: meetings that cross organizations, where nobody can see anyone else’s calendar. A poll creator picks candidate dates or a time window, shares a link, and respondents mark their availability without creating an account. Each person sees times in their own time zone, and the tool surfaces a suggested best-compromise slot once responses are in.

For recurring one-to-one scheduling, WhenFree also offers booking pages with buffer times, availability rules, and automatic video-call link creation. Anyone who connects a calendar, Google, Outlook, Apple, or iCal, gets their busy and free blocks factored in automatically; WhenFree reads only that free and busy status, never event titles, descriptions, locations, or attendees.

Sector Intel, an add-on tier, is a browsable repository of public agency meetings kept current by scheduled jobs rather than manual entry. Coverage today spans U.S. fisheries management, federal regional councils, interstate marine fisheries commissions, RFMOs and treaty bodies, tribal co-management bodies, and state agencies, with energy and transportation regulatory bodies live and expanding. A user with a connected calendar can check a proposed meeting time against this public-meeting feed directly, so scheduling conflicts with agency business are visible before they happen.

The core scheduling functions, polls, booking pages, calendar sync, are free without limits; Sector Intel is a paid layer on top, used by consultants, agency staff, and other professionals who coordinate across organizational lines. WhenFree also operates at its own established address, whenfree.io.

SpeciesID Pro reads PDFs, scanned field guides and monographs, spreadsheets, and text files, up to 500MB per upload. Scanned documents that don’t yield clean extracted text fall back to an AI vision pass that reads the page images directly, rather than failing outright.

The document is split into chunks and each chunk is sent through an extraction pass that pulls whichever fields the user has selected, from core taxonomy (scientific name, author, family, order) to physical description, ecology, and conservation status. Every extracted species entry comes back with a confidence score and a verbatim quote from the source document, so a reviewer can trace each claim back to its page.

Extracted names are then checked live against ten taxonomic databases: WoRMS, GBIF, OBIS, Catalogue of Life, ITIS, NCBI Taxonomy, Encyclopedia of Life, Plants of the World Online, World Flora Online, and BOLD Systems, weighted by the declared domain (marine, freshwater, terrestrial, or all). Synonyms resolve to their accepted name, authority, and rank, and each match links to its source record so the citation is checkable rather than asserted.

Results export as CSV, JSON, Excel, PDF, or Darwin Core, the last formatted for direct submission to GBIF or OBIS. The tool is built for environmental consultants processing survey reports, academic researchers assembling citation-ready species lists, and agencies and NGOs converting legacy documents into structured biodiversity data. What separates it from asking a general chatbot to identify species is that every claim is checked against a live registry rather than recalled from a language model’s training data; a standing disclaimer on the product itself is direct that AI-extracted results should be reviewed by a qualified taxonomist before publication.

EPN runs eight source adapters across three ingestion modes. Federal Register and regulations.gov are queried live, in real time, and also polled hourly to keep the browse feed warm. Congress.gov, CourtListener, FEC, OpenStates, and LegiScan are polled on a fixed schedule and stored, since their APIs either lack usable free-text search or carry rate limits too tight to query live. Municipal town boards, currently Westport, Massachusetts’s Conservation Commission, Planning Board, and Board of Health, are ingested by scraping HTML and PDF agendas on a schedule, since no API exists for a municipal body. The stored corpus currently holds roughly 180,000 records.

On top of the corpus sits a tool-using AI chat workspace: the model queries the stored records, hits live-search sources directly, and pulls campaign-finance docket data from a companion service, returning answers with citations back to the underlying filed document. Because municipal filings are frequently scanned PDFs with no text layer, a vision pipeline rasterizes candidate pages and routes only the ones that need full extraction to the primary model, controlling cost.

The product is aimed at environmental nonprofits, pro bono attorneys, and, through a free tier, members of the public who want to track a comment period, bill, or piece of litigation in plain language.

What differentiates EPN from federal- and state-only trackers is the municipal layer: conservation commissions, planning boards, and boards of health, parsed at the level of the filed document rather than the agenda line. Today that layer is proven with one working pilot, Westport, Massachusetts. The product is early: only one municipality is covered, there are no named partner deployments, and several outward-facing details, including a scheduling link and contact information, are still explicit placeholders pending sign-off.

TideFrame ingests dated aerial and satellite imagery for a user-drawn area of interest, drawing on state orthoimagery programs (New York, Maryland, Maine, Texas, and Washington so far), USDA NAIP, and Copernicus Sentinel-2, and reprojects every scene to the same site bounds so frames from different years and providers line up exactly. Where survey coverage exists, it overlays the published record of habitat and regulatory layers: submerged aquatic vegetation, salt marsh, wetlands, historical shoreline change, hydrography, oyster reef, aquaculture leases, permits, and outfalls.

Where no authoritative survey exists, TideFrame can derive a shoreline directly from the imagery: it computes a water index on each year’s satellite scene, thresholds it to separate water from land, vectorizes the boundary, and fits a trend across years to produce a shoreline-change rate. Because a tide difference between two scenes can look like meters of shoreline movement on a flat beach, detected waterlines are corrected to a common datum using NOAA water-level data, with the correction’s uncertainty carried through rather than hidden. Tide-station coverage today is curated to two demonstration geographies, the Florida Gulf and New York’s Atlantic coast, not nationwide.

The tool includes on-image measurement, a blind-comparison workflow where a second reviewer independently digitizes a site so two boundaries can be checked for agreement, and gated, time-limited link sharing for handing a site to a colleague or funder. It is built for restoration scientists, environmental consultants, and agency staff who need a timeline they can cite and defend, not just a picture that looks aligned. It is explicit about the limits of passive optical imagery: it will not detect submerged vegetation in turbid water, and the interface distinguishes “not detected” from “not present,” and “not surveyed” from “no habitat,” rather than papering over a data gap.

This is the newest tool in the suite and the least finished for outside viewing: its public marketing page is not currently rendering correctly. The underlying analysis pipeline is real and running on a production schedule; the public presentation layer is not yet ready for an outside audience.

Reel Boards runs the live leaderboard for a fishing tournament from the weigh station outward. A director logs a catch, weight or release, from any browser, and the standings recalculate and push out to every open leaderboard, kiosk display, and embedded widget at once. There is no app to install and no spreadsheet to maintain during the event.

Each tournament or season gets its own branded page under a unique URL, with the organizer’s logo, colors, and scoring rules. The platform is built for three groups: independent tournament directors running a single event, fishing clubs running member tournaments, and series organizers running multi-stop circuits that need cumulative standings across the season. A seven-event offshore billfish circuit runs its live leaderboard on the platform today.

Scoring supports weight-based, release-based, and points-based categories in the same tournament, plus custom divisions by species, boat type, or angler class. Multi-event series use best-of-N scoring with participation bonuses and configurable tie-breakers, so a season standing reflects a director’s specific rules rather than one fixed formula.

Results can be entered one at a time or imported in bulk from a spreadsheet, and standings export for awards and record keeping. A leaderboard can be shared as a link, displayed in kiosk mode at the dock, embedded in a tournament website, or generated as a share-ready image for social media.