Analisa US API

The server process answers. Never touches the database: use it for container health checks.

Never cached.

curl --compressed 'https://tryanalisa.com/api/health'

The database answers a trivial query within 1 s (200), else 503 with Retry-After. For load balancers and deploy checks.

Never cached.

curl --compressed 'https://tryanalisa.com/api/ready'

OpenAPI 3.1, generated from the typed registry the routes are documented in (apps/analisa/src/lib/api-docs).

Cached for a short time (see Cache-Control).

curl --compressed 'https://tryanalisa.com/api/v1/openapi.json'

Every category, indicator, source and scoring profile, the latest loaded release of each source, and the current dataset version with the versioned dataset URLs.

Cached for a short time (see Cache-Control).

curl --compressed 'https://tryanalisa.com/api/v1/meta'

The map dataset of one level (state, metro, county or city): for every place, the latest value, percentile score, period and data-quality flags of every indicator. Scores rank a place only among places of its own level, so scores of different levels are not comparable. Cities are the Census places with at least 1,000 residents (smaller ones are stored but not scored). A borrowed flag marks a city's value that is really its county's; an aggregated flag a metro area's or state's value rolled up from counties (see Places, levels and GEOIDs). Binary by default (about 630 KB with brotli for the 3,144 counties); ?format=json for JSON, whose places columns are geoid, name, stateCode, lat and lng (the county and metro columns are in the binary header).

Cached per dataset version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from GET /api/v1/meta (version) or the X-Dataset-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/datasets/state?format=json'

Every published period of one indicator for every place of a level (state, metro, county or city). Places follow the order of the matching dataset.

Cached per dataset version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from GET /api/v1/meta (version) or the X-Dataset-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/indicators/median_household_income/series?level=state'

The nation, a state, a metro area (M + CBSA code), a county or a city or town (7-digit place GEOID) with every published value of every indicator (about 20 KB with brotli) and links to the areas it belongs to or overlaps (states, metro, counties with residents' shares).

Cached per dataset version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from GET /api/v1/meta (version) or the X-Dataset-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/48453'

Finds states, metro and micro areas, counties, and cities and towns (about 36,000 places) by name. q is a name, a prefix or a misspelling ("aus", "austn"), optionally followed by a state ("austin, tx", "austin texas"); a comma makes the state certain, while without one the state is dropped when it leaves no result ("Mount Pleasant in" is not a place in Indiana). Fewer than 2 letters or digits answer an empty list. Results are ordered exact name, name prefix, word prefix, then similar spelling, the larger place first within a tier. Each has its GEOID, label, population and page. Identical queries are answered from a cache per data version; a burst of distinct queries is shed with 503 and Retry-After.

Cached per places version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/meta (version, the dataset version that search results follow) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/search?q=austin%2C%20tx&limit=5'

The counties whose census tracts make up the neighborhoods of a county, a city or a metro area, each with the content hash of its tract shapes, so a client can fetch every county's tract dataset and shapes in parallel. A city spans whole counties here: its neighborhoods are all the tracts of the counties its residents live in.

Cached per tracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from version of GET /api/v1/datasets/tract/areas/{geoid} or the X-Data-Version header of a neighborhoods response; it follows the dataset version) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/datasets/tract/areas/4805000'

The census tracts ("neighborhoods", about 4,000 residents each) of one county, in the format of /api/v1/datasets/{level}: the latest value, score, period and data-quality flags of every indicator with a tract figure (ACS 5-year estimates with their unreliable flags, CDC PLACES crude prevalence flagged crude and model-based, FEMA's hazard percentile among tracts, distances), plus the county's own value of area-wide measures, flagged borrowed. Scores rank a tract among every tract of the nation (reliability-adjusted toward its county's mean where a standard error is published), so they compare across counties; tracts under 500 residents carry values but no scores. Indicators with no value in the county are left out. Binary by default (Travis County: 37 KB with brotli; Los Angeles County, the largest: 266 KB); ?format=json for JSON.

Cached per tracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from version of GET /api/v1/datasets/tract/areas/{geoid} or the X-Data-Version header of a neighborhoods response; it follows the dataset version) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/datasets/tract?county=48453&format=json'

One census tract's own published estimates with their standard errors and 90% margins of error (ACS: published margin; CDC PLACES: from the confidence interval). Tract estimates are noisy: a difference between two tracts smaller than their margins is not a difference. The tract dataset leaves standard errors out to stay small; this is what to show next to a selected tract's values.

Cached per tracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from version of GET /api/v1/datasets/tract/areas/{geoid} or the X-Data-Version header of a neighborhoods response; it follows the dataset version) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/datasets/tract/estimates/48453001101'

Boundaries too numerous for one file, as TopoJSON with the feature id = the GEOID and properties.name = its short name. tracts/{county}: the census tracts of one county (object tracts). cities/{state}: the cities and towns of one state (object cities). 1:500,000 cartographic boundaries, simplified for map zoom and clipped to the shoreline (for maps, not for measuring areas). Addressed by content: with h set to the current hash (shard of /api/v1/datasets/tract/areas/{geoid}, or shards of the tract and city datasets) the response is cached as immutable for a year; without it, or with an outdated hash, it is the current shapes cached for five minutes. The ETag is the content hash.

Cached for a short time (see Cache-Control).

curl --compressed 'https://tryanalisa.com/api/v1/geo/shards/tracts/48453'

Household income brackets, age and sex, race and Hispanic origin, owners and renters, rent, home values, commute mode and time, education, age and size of homes: for the latest American Community Survey 5-year window and the one before it, each bucket with its count, 90% margin of error, share and change in share since the earlier window, the state's and the nation's distribution, and the data.census.gov table it comes from. Buckets sum published Census cells; the grouping is ours. A bucket with a margin of error above 30% of its estimate is flagged lowReliability; a change is significant only beyond the combined margin of error.

Cached per breakdowns version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any breakdowns response; it changes when a new ACS release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/4805000/breakdowns'

SOC 2018 occupations whose title or code matches q, best first (an exact title, then titles whose words start with the query, then titles containing it; detailed occupations before group totals; larger occupations first). Without q, the largest occupations nationally. Use a returned code with /api/v1/places/{geoid}/wages.

Cached per wages version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any occupations or wages response; it changes when a new BLS release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/occupations?q=nurse&limit=5'

Employment and annual pay (mean, median, 10th/25th/75th/90th percentiles, location quotient) by occupation from the BLS Occupational Employment and Wage Statistics survey, all industries and ownerships, next to the principal state's and the nation's. BLS publishes metro areas, states and the nation: those answer as themselves, a county or city with the metro area it lies in (requested says which place was asked about), and a place in no metro area is a 404. With occupation the one occupation; otherwise the largest occupations by employment (group, limit). Pay is in current dollars, not adjusted for the cost of living, and BLS caps the highest wages it publishes (wage-top-coded).

Cached per wages version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any occupations or wages response; it changes when a new BLS release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/M12420/wages?occupation=15-1252'

For every scored indicator with a comparable history: the latest value and the one about five years earlier (two non-overlapping periods of the same publisher dataset; for American Community Survey indicators the windows 2015-2019 and 2020-2024), the change, whether it is an improvement for the indicator's direction, whether it exceeds the published margin of error (1.645 × the root of the summed squared standard errors; indicators without a margin are unassessed), and how favourable it is against the same change in the other places of the level (peers). Also the five largest improvements and declines, and up to five places that looked most like this one five years ago and have improved most since (most real improvements minus declines, then the rise in standing). Never compared across a source change. A comparison of published estimates: it says what moved, not why. Cities under 1,000 residents are not ranked and answer 404.

Cached per changes version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any changes response; it changes when a release of any indicator source is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/48453/changes'

Employer establishments, paid employees, annual payroll and nonemployer businesses for the NAICS sectors and a curated list of narrower industries (restaurants, child day care, dentists, gyms, auto repair…), from the Census Bureau's County Business Patterns (2023) and Nonemployer Statistics (2024), with establishments per 10,000 residents against the state's and the nation's, a density index (1.00 = the nation's density) and average pay. Metro areas are the exact sums of their counties (flag aggregated). The Census Bureau publishes business counts no finer than counties, so a city answers with its county's figures (requested names the city). Employment and payroll are noise-infused by the Census Bureau (see flags). Counts of businesses against counts of residents: not a measure of demand, size, quality or profit.

Cached per business version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any business response; it changes when a new County Business Patterns or Nonemployer Statistics release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/places/48453/business'

The all-industries total, the 19 NAICS sectors and the curated narrower industries, each with its stable key, its NAICS code in both editions the surveys use (County Business Patterns 2023 is NAICS 2017, Nonemployer Statistics 2024 is NAICS 2022) and national establishments, employment, payroll and nonemployers. Use a key or NAICS code with /api/v1/business/industries/{naics}/places.

Cached per business version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any business response; it changes when a new County Business Patterns or Nonemployer Statistics release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/business/industries'

Places ranked for one industry: establishments, employment, average pay, nonemployers, establishments per 10,000 residents, residents per establishment (more residents per business means fewer competitors per resident: a count, not a measure of demand), a density index against the nation's and, once the tables hold a second reference year of the same NAICS edition, growth of establishments (only 2023 is loaded now, so growth is null). With federal contract obligations in the industry's NAICS codes for the contracts fiscal year: the nation's, and the ten metro areas (sums of their counties), counties and cities (estimated from ZIP codes) where the most was performed. Metro areas count under their principal state. Places below minPopulation or with fewer than minEstablishments are left out: a handful of businesses makes the ratios jump around.

Cached per business version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any business response; it changes when a new County Business Patterns or Nonemployer Statistics release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/business/industries/restaurants/places?state=TX&limit=10'

Every period × every counterpart country, plus the publisher's all-countries total, for the nation or a state (10–100 KB).

Cached per migrations version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from the X-Dataset-Version header of any migrations response) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/migrations/foreign_born_population?hub=US'

International migration against natural increase, the share born abroad and long national totals.

Cached per migrations version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from the X-Dataset-Version header of any migrations response) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/migrations/context'

The loaded fiscal years, the agency filter options, the latest action date loaded and the contracts data version.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/meta'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/map?level=state&metric=per_capita'

KPIs against the previous year, monthly and yearly series, categories, top buyers, recipients and industries, largest awards. Place = place of performance. A metro area's figures are the sum of its counties'. A city's are estimated: each action's county money split by the share of its award's ZIP code that lies in the city (land, Census ZCTA relationship files), part of its counties' figures (see attribution); counts are of actions with any share in the city.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/places/06?fy=2025'

Trigram and full-text matching; 3 normalized characters or more (shorter queries answer empty groups). Also organizations with federal assistance and no contracts, and Assistance Listings by number or title.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/search?q=lockheed'

Keyset pagination: pass nextCursor back as cursor. An unknown fiscal year or agency, or a malformed cursor, is a 400; other unrecognized filters are ignored.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/awards?fy=2025&place=48&sort=value'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/awards/8469442'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/recipients?fy=2025'

The response version combines the data and company-family versions: a URL pinned to the data version alone is revalidated by ETag rather than cached as immutable.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/recipients/1451334?fy=2025'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/agencies?fy=2025&place=06'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/agencies/8838?fy=2025'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/opportunities?open=1&sort=deadline'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/opportunities/9832ec694bd8461bb279fd5f19e43e5a'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/graph?fy=2025&limit=80'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/relationship?subtier=8917&recipient=2999810&fy=2026'

Every integrity signal for the scope with the national value, top contributors and the sole-source justifications cited.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/signals/US?fy=2025'

Agencies, offices, recipients or relationships by obligations or an integrity measure, with a minimum size for shares.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/rankings?kind=agencies&metric=not_competed_share'

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/flows/US?mode=agency-category'

Contract awards whose current period of performance ends between today and the window's end, and whose potential value (base plus every option) is at least the floor, soonest end date first, 25 per page (keyset pagination: pass nextCursor back as cursor). Each row says whether options remain (an option may be exercised instead of recompeting), whether the award was sole-source, who the incumbent is, and up to three SAM.gov notices that may relate to it (match: same solicitation, cites the contract number, or an open notice of the same office for the same NAICS or PSC code, which is only a hint). A recompete candidate is not a prediction: the data cannot tell whether the agency will compete, extend or end the work. End dates are those of the latest loaded modification (lastActionDate). With summary=1 the first page also carries the totals of everything selected, by month and by toptier agency. Anonymous and free callers get the window and the value floor, the soonest ten awards and the totals; one buyer (agency, subtier, office) or one company (recipient) shows five rows to everyone, as profiles do. The recompete calendar plan (Pro and above) gets every row and every filter, for a signed-in session or an API key of its owner (Authorization: Bearer); the CSV export needs the session. access and filters in the payload say what applied; access.ignored lists the filters a preview dropped. The answer depends on the caller's plan (session or key), so it is private and Vary: Cookie. An unknown window, value, agency or place, a malformed filter or cursor is a 400.

Cached for a short time (see Cache-Control).

curl --compressed 'https://tryanalisa.com/api/v1/contracts/recompetes?window=12&min=10000000&summary=1'

Companies that win from the same buyers in the same markets as the company (the recipient's whole company family). A market is a subtier agency × NAICS industry group (first four digits of the award's code); the company's obligations by market over the three fiscal years ending in fy are compared with every other company's by weighted Jaccard similarity (similarity, 0 to 1), so a company of the same size and mix ranks above a giant that merely also sells there. The 300 largest markets of the company are compared and the 500 companies sharing the most dollars are ranked; organizations only. Each competitor carries its obligations by year, the percent change between the last two complete years and its top three shared markets. Prime contract obligations only: subawards are never added. Everyone gets the top three; a plan that includes the feature gets the full lists. The payload is built for the caller's plan (access says which: rows beyond limit are never sent), so the response depends on the session: it is Vary: Cookie, and for a caller with the plan it is private. format=csv downloads the full list (UTF-8 with BOM; 401 signed out, 402 or 403 without the plan). 404 for an unknown recipient.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/recipients/1461780/competitors?fy=2026'

The loaded fiscal years, the agency filter options (assistance agency ids), the date the data is current to and the assistance data version.

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/meta'

Obligations, obligations per resident, or loan face value (face_value) by place of performance. Unknown level or metric values fall back to county and obligations.

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/map?level=state&metric=per_capita'

KPIs against the previous year, monthly and yearly series, types, programs, top agencies and top organizations. County aggregates and individuals are counted in the totals (aggregateShare, individualShare), never listed. A metro area's figures are the sum of its counties'.

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/places/06?fy=2025&type=grants'

404 when it received none in the loaded years (or it is not an organization: individuals and county aggregates have no profile).

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/recipients/3002653?fy=2025'

A toptier agency (its subtiers together) or a subtier agency. 404 for offices and for agencies that awarded no assistance.

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/agencies/8933?fy=2025'

Programs of a fiscal year, optionally of one type group and matching q. With type=loans the ranking and shares follow face value (loan obligations are subsidy costs).

Cached per assistance version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/assistance/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/assistance/programs?fy=2025&q=medicaid'

The companies a company reported paying as a prime (subcontractors), the primes that reported paying it (primes), and potential partners: sub-recipients that the company's competitors (see the competitors route) reported paying and that have no report with the company in either direction, ranked by how many competitors paid them. The company is the recipient's whole family (every entity's UEI); a partner that is a contract recipient is its whole family too. Every loaded fiscal year is counted; each partner lists its fiscal years and, for sub-contracts, the industries and buyers of the prime awards behind it. Reported amounts are first-tier subawards primes report themselves (FFATA; many go unreported), money already inside the primes' obligations: never add them to prime totals. Organizations only. Everyone gets the top three; a plan that includes the feature gets the full lists. The payload is built for the caller's plan (access says which: rows beyond limit are never sent), so the response depends on the session: it is Vary: Cookie, and for a caller with the plan it is private. Potential partners and format=csv (the full lists in one table) are part of the plan (401 signed out, 402 or 403 without it). 404 for an unknown recipient.

Cached per subawards version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of a subawards response, which combines the contracts and subawards versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/subawards/recipients/1461780/teaming'

Totals, the largest sub-recipients and the largest subawards the prime reported, every loaded year. reported is null when it reported none; 404 for an unknown award.

Cached per subawards version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of a subawards response, which combines the contracts and subawards versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/subawards/awards/18603297'

The sub-recipients a contract recipient reported paying and the primes that reported paying it (matched by UEI), for a fiscal year, with both amounts by year. 404 for an unknown recipient.

Cached per subawards version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of a subawards response, which combines the contracts and subawards versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/subawards/recipients/1458732?fy=2026'

Totals by kind and the largest sub-recipients under a toptier or subtier agency's prime awards in a fiscal year. 404 for offices and unknown agencies.

Cached per subawards version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of a subawards response, which combines the contracts and subawards versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/subawards/agencies/8838?fy=2026'

Its SAM.gov exclusion records and its awards with actions dated during one; every loaded year (exclusions span years). exclusions are exact identifier matches (reported); possible are name matches an AI judged likely (shown, never counted). Only exclusions that bar awards count, and only new awards, options and added work. 404 when it has none.

Cached per exclusions version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an exclusions response, which combines the contracts and exclusions versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/exclusions/recipients/1461087'

The recipient's exclusions and the award's actions dated while one was in effect. exclusions are exact identifier matches (reported); possible are name matches an AI judged likely (shown, never counted). Only exclusions that bar awards count, and only new awards, options and added work. 404 when the recipient has no exclusion.

Cached per exclusions version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an exclusions response, which combines the contracts and exclusions versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/exclusions/awards/37562989'

Counted actions of a toptier, subtier or office in a fiscal year and in every loaded year, the recipients and the awards (exact identifier matches only). 404 when there are none.

Cached per exclusions version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an exclusions response, which combines the contracts and exclusions versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/exclusions/agencies/8926?fy=2024'

Lobbying disclosures (House Clerk: LD-2 reports, LD-203 contributions) and PAC activity (FEC) of the organizations matched to the recipient's company family, every loaded year and cycle. Organizations are matched to contract recipients by AI name judgments (matches, ai): the links are inferred, and a match shows a correlation, never influence. Empty matches (and null lobbying, pacs) when nothing matched; 404 for an unknown recipient.

Cached per influence version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an influence response, which combines the contracts, company-family and influence versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/influence/recipients/1461780'

Through the organizations matched to the recipient's company family: members of Congress whose campaigns received contributions from their PACs in the latest two FEC cycles (amount and number of PACs; members' current or latest seat), and the bills and resolutions their current lobbying reports named in issue descriptions (each reference resolved to a Congress from the report's period, explicit markers and the bill's title; the description itself is not stored). Comparisons, not causes: nothing says why a contribution was made, how a member voted or which side a report took. Organizations are matched to contract recipients by AI name judgments (matches, ai): the links are inferred, and a match shows a correlation, never influence. Versioned on both the influence and the Congress data. 404 for an unknown recipient.

Cached per influence version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an influence response, which combines the contracts, company-family and influence versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/influence/recipients/1455292/congress'

Organizations whose lobbying reports list a toptier or subtier agency, for the calendar year named like the fiscal year (else the nearest loaded one), ranked by spend. A report's amount covers all its issues and agencies (reports do not split it). For matched organizations, the share of their family's contract obligations in the fiscal year that came from the agency. Organizations are matched to contract recipients by AI name judgments (matches, ai): the links are inferred, and a match shows a correlation, never influence. Clients that may be people are counted in unnamed, never named. 404 for offices and agencies lobbying reports do not name.

Cached per influence version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header of an influence response, which combines the contracts, company-family and influence versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/influence/agencies/8838?fy=2025'

The money domains, the outcomes, every domain–outcome pairing with its rationale, lag, caveat and finding (strength, shocks, interval, placebo), the method's parameters, coverage (fiscal years, shock counts) and the caveats to show.

Cached per outcomes version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/outcomes/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/outcomes/meta'

Money per resident by domain and fiscal year (with the nation's), the related outcomes by calendar year (with the nation's, and their change since the base year), and spending shocks: a county's own, each with its matched comparison counties and event study per outcome; a state's counties' (summaries). 404 for an unknown place.

Cached per outcomes version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/outcomes/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/outcomes/places/16001'

For one domain–outcome pairing of the crosswalk: the pooled event study (mean difference in change between shock counties and their comparison counties by relative year, 95% bootstrap interval clustered by county, number of shocks), the placebo, the finding and the shocks behind it. 400 for a pair outside the crosswalk.

Cached per outcomes version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/outcomes/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/outcomes/effects?domain=contracts&outcome=job_growth_rate'

The loaded Congresses (dates, district plan, roll calls per chamber, bills, laws, last vote, which is current), the next federal general election, the plan and fiscal years of contract money by district, the content hash of each plan's district shapes, and the sources. Its version pins the Congress routes that follow only Congress data.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/meta'

One entry per member and seat during the Congress: party and caucus, seat, dates served, whether seated on the reference date, and activity (roll calls held while serving, missed, party-unity votes and those cast with the caucus majority, DW-NOMINATE first dimension, bills sponsored and enacted, cosponsorships across the aisle), with the vacancies and the chamber's roll-call counts as denominators. Volume and attendance, never a grade. House positions come from the House Clerk; Senate positions and DW-NOMINATE scores from Voteview (senate.gov blocks automated access), cited as Lewis, Poole, Rosenthal, Boche, Rudkin & Sonnet, Voteview: Congressional Roll-Call Votes Database.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/chamber?congress=119&chamber=senate'

Roll calls of a Congress, newest first, 50 per page (pass next back as cursor), with totals and party tallies. Recorded roll calls only: voice votes and unanimous consent record no positions, and none is inferred. House positions come from the House Clerk; Senate positions and DW-NOMINATE scores from Voteview (senate.gov blocks automated access), cited as Lewis, Poole, Rosenthal, Boche, Rudkin & Sonnet, Voteview: Congressional Roll-Call Votes Database.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/votes?congress=119&chamber=house&category=passage'

Every member's position as recorded (yea, nay, present, not voting, or a name in Speaker elections), party tallies, which Democrats and Republicans voted against their party's majority, the measure and the other roll calls on it. sourceId names the record the positions come from. House positions come from the House Clerk; Senate positions and DW-NOMINATE scores from Voteview (senate.gov blocks automated access), cited as Lewis, Poole, Rosenthal, Boche, Rudkin & Sonnet, Voteview: Congressional Roll-Call Votes Database. 404 for an unknown roll call.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/votes/h119-1-145'

Bills and resolutions of a Congress, latest action first (by relevance for a text search), 50 per page with the total (pass next back as cursor). Stage is the furthest reached, from the Library of Congress action codes; the policy area is the one CRS assigned.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/bills?congress=119&stage=law'

Titles, the latest CRS summary, subjects, committees, the action history, cosponsors (original, withdrawn), the roll calls on it, the toptier agencies under its committees' jurisdiction (Analisa's mapping), and the organizations that named it in lobbying reports, each with its AI match to a contract recipient and the match probability: disclosed activity, never evidence of influence. 404 for an unknown bill.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/bills/119-hr-1'

Members by name (sitting members first) and bills by number (any loaded Congress) or by words (full text, the current Congress first). Names and words need at least 3 characters; a bill number may be shorter.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/search?q=water%20resources'

Roll calls by category and pass rate, votes per day, the monthly share of party-unity votes, how often both parties' majorities voted the same way, the members with the most party-unity votes against their caucus, the closest votes, bills by furthest stage reached and laws by policy area. Recorded roll calls only: voice votes and unanimous consent record no positions, and none is inferred.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/patterns?congress=119&chamber=senate'

Terms, contact, committee seats, recorded activity per loaded Congress (roll calls, missed votes, party-unity votes with the caucus, rank among members with at least 100 roll calls), DW-NOMINATE placement, the latest passage and amendment votes and breaks with the caucus, sponsored bills and identifiers (GovTrack, OpenSecrets, Wikidata, Ballotpedia, FEC). Volume and attendance, never a grade. House positions come from the House Clerk; Senate positions and DW-NOMINATE scores from Voteview (senate.gov blocks automated access), cited as Lewis, Poole, Rosenthal, Boche, Rudkin & Sonnet, Voteview: Congressional Roll-Call Votes Database. 404 for an unknown member.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/members/S001231'

Per loaded FEC cycle: contributions the member's campaigns received from political committees (direct and in-kind, by kind of giving committee), the part received from PACs whose sponsoring organization is matched by AI to a federal contract recipient (each match with its probability; the largest ten PACs beside that recipient's contract obligations), and the FEC summaries of the member's candidacies. Comparisons, not causes: money given and received is shown next to contracts, never why it was given or what followed. 404 for an unknown member.

Cached per congress-money version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a response of the same route, which combines the Congress version with the contracts version (committees, districts, district money) or the influence version (a member's money)) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/members/S001231/money'

Current members as published (title, side, rank), subcommittees, the toptier agencies in its jurisdiction (Analisa's mapping) with their contract obligations in the latest closed fiscal year, and the bills referred to it in the current Congress. 404 for an unknown committee.

Cached per congress-money version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a response of the same route, which combines the Congress version with the contracts version (committees, districts, district money) or the influence version (a member's money)) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/committees/HSAP'

Residents, land area, counties (shares both ways), the members who held the seat while the plan was in force, the state's senators, and the contract obligations attributed to the district by place of performance (weighted by its share of each ZIP code and county: an estimate). 404 when the plan has no such district.

Cached per congress-money version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a response of the same route, which combines the Congress version with the contracts version (committees, districts, district money) or the influence version (a member's money)) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/districts/cd119/CA-12'

Contract obligations attributed to every district of a plan in one fiscal year, and per resident: an estimate, by the district's share of each ZIP code and county. Both parameters are required.

Cached per congress-money version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a response of the same route, which combines the Congress version with the contracts version (committees, districts, district money) or the influence version (a member's money)) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/congress/money?plan=cd119&fy=2025'

Every House district and Senate seat of the next general election's cycle: the candidates listed (those whose campaigns received at least $5,000, and incumbents), up to three leaders by money received with their FEC summaries and the part received from PACs whose sponsor is matched by AI to a federal contract recipient, and the sitting member. Candidates are those who filed with the FEC for the seat in the cycle, including those who lost a primary or withdrew: never the ballot, which comes from the state election office.

Cached per elections version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a races response, which combines the Congress, influence and contracts versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/elections/races'

Every listed candidate and the money their campaign received, the sitting member's (for a Senate race, both senators') recorded activity in the current Congress, and for a House race the federal money in the district under the current plan (contracts attributed to it; assistance programs to its counties weighted by its share of their residents: estimates) and its counties. Candidates are those who filed with the FEC for the seat in the cycle, including those who lost a primary or withdrew: never the ballot, which comes from the state election office. 404 for a seat with neither candidates nor a sitting member.

Cached per elections version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of a races response, which combines the Congress, influence and contracts versions) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/elections/races/ca-12'

Up to 6 contested final-passage roll calls per chamber of the current Congress in the chosen policy areas (bills and joint resolutions with a CRS summary, yeas between 15% and 85% of those voting, widened to any opposition when too few), areas taken in turn: the same set for everyone with the same district and areas. Each with its CRS summary excerpt, each party's majority side, and the recorded positions of the district's representative and its state's senators. Recorded votes only.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/elections/record?district=CA-12&areas=Health,Taxation'

The congressional districts of a county or city under the current plan, with the share of its residents in each (a city through its counties: an estimate where it spans district lines). Empty when none is known.

Cached per congress version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/congress/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/elections/districts?geoid=06001'

Mail ballots transmitted, returned, counted and rejected (with the reasons reported and their shares) and provisional ballots cast, counted and rejected in each general election, from the EAC Election Administration and Voting Survey, for a county, a state or the nation, with its state and the nation for comparison. Aggregates only: no voter is identified. Counts are null where offices did not report them. 404 for an unknown place.

Cached per ballots version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from the X-Data-Version header (or version) of any ballots response; it changes when a new EAC EAVS release is loaded) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/elections/ballots/06001'

Checks up to 100 suppliers by UEI and/or CAGE code against the SAM.gov exclusions, synchronously, with the function the bulk screening of the account area uses. A result is active when a record that lists the supplier's UEI or CAGE code is currently listed, ended when only records SAM.gov no longer lists do, and none otherwise; each reported record carries its excluding agency, type, dates and SAM.gov link, and whether it bars federal contract awards. Suppliers with a UEI found among the federal contract recipients also get their obligations of the last three fiscal years, the awards dated while they were excluded, and whether they have lobbying disclosures (AI-matched). Only identifiers match: a name never finds an exclusion here (the account's list screening searches names and labels them possible matches). Results reflect the exclusions as SAM.gov published them at the vintage in sources: always confirm in SAM.gov before making a decision. Part of the supplier screening plan: send an API key (Team plan) or call it signed in (Pro); each supplier counts one row of the monthly allowance shared with uploads, and a call that does not fit is refused whole with 402. Send Content-Type: application/json (415 otherwise). Never cached.

Never cached.

curl -X POST 'https://tryanalisa.com/api/v1/screening/check' \
  -H "Authorization: Bearer $ANALISA_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"suppliers":[{"uei":"Q54JKCJTAP64"},{"cage":"3HS96"},{"uei":"ZZZZZZZZZZZZ","cage":"1A2B3"}]}'

Ranks the cities, metro areas or counties of the datasets for a person's criteria: where they want to live, household, housing budget, priorities, climate, a commute destination and must-haves. Pure formulas, no model: hard filters first (each removal counted by reason), then the weighted quality of life on the person's priorities (the catalog's two-level aggregation), a climate preference as one more category, penalties for housing over budget (capped, so an otherwise ideal place stays visible and flagged) and at most perMetro results per metro area (with a commute destination the cap is lifted to limit, because every commutable place sits in one or two metro areas). Housing costs are the typical asking rent, or the mortgage payment at the current Freddie Mac rate plus property tax; commutes are straight-line estimates, not routes. Every figure's source, vintage and period are in provenance. Anonymous calls are limited to 60 a minute per client; with an API key, to 600 a minute besides the plan's daily allowance. Send Content-Type: application/json (415 otherwise).

Never cached.

curl -X POST 'https://tryanalisa.com/api/v1/where-to-live' \
  -H 'Content-Type: application/json' \
  -d '{"level":"city","states":["TX","CO"],"sizes":["mid-size"],"housing":{"mode":"rent","monthlyBudget":2200},"household":{"adults":2,"preschoolers":1},"priorities":{"safety":5,"education":4},"climate":"mild","limit":10}'

The same ranking for the compact criteria the /where-to-live page puts in its links: c is the base64url of the criteria JSON with defaults left out. An unreadable value gives the defaults, and a value with some invalid fields keeps the valid ones. Ranks the cities, metro areas or counties of the datasets for a person's criteria: where they want to live, household, housing budget, priorities, climate, a commute destination and must-haves. Pure formulas, no model: hard filters first (each removal counted by reason), then the weighted quality of life on the person's priorities (the catalog's two-level aggregation), a climate preference as one more category, penalties for housing over budget (capped, so an otherwise ideal place stays visible and flagged) and at most perMetro results per metro area (with a commute destination the cap is lifted to limit, because every commutable place sits in one or two metro areas). Housing costs are the typical asking rent, or the mortgage payment at the current Freddie Mac rate plus property tax; commutes are straight-line estimates, not routes. Every figure's source, vintage and period are in provenance.

Cached per dataset version: ETag (answer 304 to If-None-Match) and X-Dataset-Version; pin the version with v (read it from GET /api/v1/meta (version) or the X-Dataset-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/where-to-live?c=eyJ2IjoxLCJzdGF0ZXMiOlsiVFgiXX0'

The awards list with the same filters and order, as a CSV download (UTF-8 with BOM, CRLF). Streamed; one export at a time per server share (503 + Retry-After when busy). Text cells that start like a spreadsheet formula are prefixed with an apostrophe.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.

curl --compressed 'https://tryanalisa.com/api/v1/contracts/export/awards?fy=2025&place=06'

A server ranking (top 50, as /api/v1/contracts/rankings), or with kind=places every county or state on a map metric, ranked.

Cached per contracts version: ETag (answer 304 to If-None-Match) and X-Data-Version; pin the version with v (read it from GET /api/v1/contracts/meta (version) or the X-Data-Version header) for a response cached as immutable.