--- title: "Data Hippo — public website" description: "English public page copy. Generated from the website; excludes login, forms and live calendar data." canonical: "https://datahippo.ai/" lang: "en" modified: "2026-09-18T00:32:11.000Z" --- # Data Hippo — public website English public page copy. Generated from the website; excludes login, forms and live calendar data. ## Markdown navigation - [Whole site](https://datahippo.ai/site.md) - [Home](https://datahippo.ai/index.md) - [Contact](https://datahippo.ai/contact.md) - [Platform guide (English)](https://datahippo.ai/docs/index.md) - [Product guide (English)](https://datahippo.ai/docs/products.md) - [Booking guide (English)](https://datahippo.ai/docs/booking.md) - [Privacy (English)](https://datahippo.ai/privacy.md) - [Terms (English)](https://datahippo.ai/terms.md) - [BAA (English)](https://datahippo.ai/baa.md) Available languages: [English](https://datahippo.ai/site.md) · [Português](https://datahippo.ai/pt/site.md) · [Español](https://datahippo.ai/es/site.md) · [Français](https://datahippo.ai/fr/site.md) - [/](https://datahippo.ai/index.md) - [/contact/](https://datahippo.ai/contact.md) - [/docs/](https://datahippo.ai/docs/index.md) - [/docs/products/](https://datahippo.ai/docs/products.md) - [/docs/booking/](https://datahippo.ai/docs/booking.md) --- # Healthcare Data & AI Platform | Data Hippo Source: https://datahippo.ai/ Forecast radiology demand. # Healthcare Data & AI Systems for Every Decision [Book your demo today](https://datahippo.ai/contact.md) Platform ## Turn your data into _decisions you can act on._ - Connect. Bring your systems together, in your cloud. - Reconcile. Trace every number back to its source. - Decide. Give your team answers they can trust. - Act. Power workflows and AI agents with live data. - Own. Keep the foundation you build. [Book a demo](https://datahippo.ai/contact.md) ### Results you can measure. - $300K+ — Annual revenue recovered - 6 hours → 10 minutes — Reporting questions - 1 in 5 — Extra coverage shifts prevented Products ## Start with one challenge. _Build from there._ ### See what’s changing while you can still act. Track worklists, throughput, capacity, and bottlenecks as they happen. How it works Connect operational data, keep metrics current as records change, and follow worklists, throughput, and capacity in a shared live view. [See real-time](https://datahippo.ai/contact.md) ### See demand before it hits the floor. Forecast volume, backlog, and turnaround risk across sites and modalities. How it works Use historical demand to forecast volume and backlog. Compare demand with capacity across sites and modalities to inform planning. [See forecasting](https://datahippo.ai/contact.md) [Product demo: Data Hippo detects a demand surge weeks ahead and the plan adjusts to cover it](https://datahippo.ai/assets/forecasting-storyboard/forecast-planning-demo.mp4) ### Let the data watch for you. Set what matters once. Data Hippo checks it continuously and alerts the right team when something changes. How it works Define what to watch in your data. Data Hippo checks for changes and routes alerts to the team that needs to respond. [See alerts](https://datahippo.ai/contact.md) [Product demo: Data Hippo continuously watches what matters and alerts the right team when attention is needed](https://datahippo.ai/assets/forecasting-storyboard/monitor-alert-demo.mp4) ### Stop doing the same work every month. Turn recurring reports, workflows, and handoffs into governed automation. How it works Set up a recurring report or workflow. Data Hippo builds the report, checks the results, and delivers it through a governed process. [See reporting](https://datahippo.ai/contact.md) [Product demo: a governed recurring report builds, checks, and delivers itself automatically](https://datahippo.ai/assets/forecasting-storyboard/automated-work-demo.mp4) ### Ask a question. Build on the answer. Query data, engineer pipelines, and build applications with AI agents on one governed foundation. Start with a question. Keep the sources in view. How it works Ask a question in plain language. Data Hippo uses governed data to produce an answer with its sources, so your team can review the evidence. [See the agent](https://datahippo.ai/contact.md) [Product demo: a governed AI agent turns a question into a source-backed answer and action](https://datahippo.ai/assets/forecasting-storyboard/hippo-agent-v16-demo.mp4) ### What decision should your data make easier? From planning and operations to finance, automation, and AI, the platform is built around the problems your team needs to solve. [Bring us a challenge](https://datahippo.ai/contact.md) **Capacity planning** See the demand that is coming and size rooms, scanners, and shifts to it. **Leakage detection** Find the revenue that slips between order, exam, and claim. **Staffing** Put the right people on the right days before the week starts. **Decision support** Answers with their source attached, ready for the meeting. **Contract intelligence** Know what each payer contract actually pays, and where it does not. **Custom apps** Small apps on your own data, built in days inside your cloud. **Scheduling** Fill the schedule against real demand, not last year’s template. **Risk detection** Catch the anomaly before it becomes an incident. **Quality** Track turnaround, repeats, and outliers where the work happens. **AI agents & Data MCPs** Agents that work your data through governed connectors. Turn data into decisions. **Turn decisions into action.** Teams ## Built for the teams solving _healthcare’s hardest challenges._ [Talk to us about your opportunity](https://datahippo.ai/contact.md) - Operations ### Operators & COOs See what is off plan this week and act on it, fast. - Off-plan this week - Act on it - COO & CFO ready - Margin & cost - Revenue cycle ### Finance & RCM leaders Find denials, AR days, and underpayments before the month closes. - Denials & underpayments - AR days - Net collection rate - Charge capture - Technology ### Technology & data leaders A governed, secure foundation in your own cloud, with lineage you can audit. - Governed & secure - Stays in your cloud - Lineage you can audit - Scales with you - Imaging ### Radiology & imaging leaders Turn reads, turnaround, and referral patterns into staffing moves chiefs can trust. - Reads & turnaround - Referral patterns - Capacity moves - Section chiefs Security ## Security first. By design. **First, do no harm.** We built Data Hippo to be secure by design, not secured after the fact. Designed to run in your environment, under your keys and controls. Governance is enforced on every query, every answer is traceable to its source, and your data is never used to train a model. Security is simpler when it starts with the architecture. - **Runs in your environment** — By design - **BAA for eligible customers** — On demand - **Retain sovereignty over your data** — Retained - **Your keys, no backdoors** — Customer-managed - **Row and column-level governance on every query** — Enforced - **Lineage to the source row, full audit log** — Auditable - **Encryption in transit and at rest** — Protected A BAA requires written eligibility confirmation and a signed agreement. [Read the BAA terms](https://datahippo.ai/baa.md) Built on open standards for on-premises and cloud. - [AWS](https://aws.amazon.com/) - [Microsoft Azure](https://azure.microsoft.com/) - [Google Cloud](https://cloud.google.com/) - [Databricks](https://www.databricks.com/) - [Snowflake](https://www.snowflake.com/) Pricing ## Start with one problem. Keep the platform. ### Hippopotamus Your data products, run by you. A small monthly commitment - Ingestion, modeling, analytics & forecasting - Automation & AI workflows - Runs in your cloud - Guided setup when you need it For teams that run it themselves. [Book a demo](https://datahippo.ai/contact.md) ### Hippocrates Your data team, run for you. Less than a full-time data hire - Everything in Hippopotamus - Fractional data leadership - A team to build and operate it - Automation & AI, run for you For teams that want us to run it. [Book a demo](https://datahippo.ai/contact.md) ### Hippocampus Your data office, built with you. Without building one from scratch - Everything in Hippocrates - Dedicated data leadership & team - Built around your systems and needs - Custom AI, product & data solutions For enterprises that want a data office. [Talk to a founder](https://datahippo.ai/contact.md) **Start small.** Scale with confidence. TRY NOW ## Prove it on _your data_, in your cloud. Start with one real challenge and prove it in 90 days. Everything runs in your environment, under your controls, and the platform you build stays yours. 90 days Prove the first challenge. Real data. Real environment.Prove the first challengeReal data. Real environment. Yours Keep the platform. No lock-in.Keep the platformNo lock-in. 100% Traceable to source. Every answer, back to the data.Traceable to sourceEvery answer, back to the data. 0 Data leaves your cloud. Your environment. Your controls.Data leaves your cloudYour environment. Your controls. [Book your demo today](https://datahippo.ai/contact.md) --- # Book a Demo | Data Hippo Source: https://datahippo.ai/contact/ # Start the conversation. Schedule a demo or reach out with questions. You'll hear directly from the founders, usually within a few hours. [contact@datahippo.ai](mailto:contact@datahippo.ai) ## Founder-led Data Hippo demo - 30 minutes - Google Meet - Eastern Time Or email [contact@datahippo.ai](mailto:contact@datahippo.ai) --- # Healthcare data platform guide | Data Hippo Source: https://datahippo.ai/docs/ # Healthcare data platform guide A reference to Data Hippo’s public platform, deployment, security, and product information. ## What is Data Hippo? Data Hippo turns healthcare data into source-backed decisions, workflows, and AI agents, in your own cloud. Start with one challenge. Keep the platform. Start with one challenge. Keep the platform. Explore [the product guide](https://datahippo.ai/docs/products.md) for real-time operations, forecasting, monitoring, automation, and AI agents. ## How does the foundation support decisions? 1. **Connect.** Bring your systems together, in your cloud. 2. **Reconcile.** Trace every number back to its source. 3. **Decide.** Give your team answers they can trust. 4. **Act.** Power workflows and AI agents with live data. 5. **Own.** Keep the foundation you build. [Explore the platform overview →](https://datahippo.ai/index.md) ## Where does Data Hippo run? Built on open standards for on-premises and cloud. Runs in your environment by design. Production deployments are governed by separate written agreements. Deployment and controls should be discussed for the offering and environment you are evaluating. ## What security controls are described? - **Runs in your environment:** By design. - **BAA for eligible customers:** On demand. - **Retain sovereignty over your data:** Retained. - **Your keys, no backdoors:** Customer-managed. - **Row and column-level governance on every query:** Enforced. - **Lineage to the source row, full audit log:** Auditable. - **Encryption in transit and at rest:** Protected. [Security overview](https://datahippo.ai/index.md) · [Business associate agreement](https://datahippo.ai/baa.md) · [Privacy Policy](https://datahippo.ai/privacy.md) ## How do we start? Start with one real challenge and prove it in 90 days. Compare [the service options](https://datahippo.ai/index.md), then [book a founder-led demo](https://datahippo.ai/contact.md) to discuss your team’s systems and priorities. This guide describes the public website. It does not grant access to customer data, private deployments, or platform APIs. --- # Product and workflow guide | Data Hippo Source: https://datahippo.ai/docs/products/ # Product and workflow guide What each Data Hippo product demonstration shows, with readable explanations alongside the public videos. Each product is an entry point into one governed foundation. These explanations describe the workflows demonstrated on the homepage; they are not customer case studies. ## Real-time operations ### See what’s changing while you can still act. Track worklists, throughput, capacity, and bottlenecks as they happen. **How it works.** Connect operational data, keep metrics current as records change, and follow worklists, throughput, and capacity in a shared live view. [Explore the product demonstrations →](https://datahippo.ai/index.md) ## Forecast & plan ### See demand before it hits the floor. Forecast volume, backlog, and turnaround risk across sites and modalities. **How it works.** Use historical demand to forecast volume and backlog. Compare demand with capacity across sites and modalities to inform planning. [Explore the product demonstrations →](https://datahippo.ai/index.md) ## Monitor & alert ### Let the data watch for you. Set what matters once. Data Hippo checks it continuously and alerts the right team when something changes. **How it works.** Define what to watch in your data. Data Hippo checks for changes and routes alerts to the team that needs to respond. [Explore the product demonstrations →](https://datahippo.ai/index.md) ## Automate work ### Stop doing the same work every month. Turn recurring reports, workflows, and handoffs into governed automation. **How it works.** Set up a recurring report or workflow. Data Hippo builds the report, checks the results, and delivers it through a governed process. [Explore the product demonstrations →](https://datahippo.ai/index.md) ## Ask your data ### Ask a question. Build on the answer. Query data, engineer pipelines, and build applications with AI agents on one governed foundation. Start with a question. Keep the sources in view. **How it works.** Ask a question in plain language. Data Hippo uses governed data to produce an answer with its sources, so your team can review the evidence. [Explore the product demonstrations →](https://datahippo.ai/index.md) ## What else can the platform support? From planning and operations to finance, automation, and AI, the platform is built around the problems your team needs to solve. The homepage also highlights capacity planning, staffing, contract intelligence, scheduling, quality, leakage detection, decision support, custom apps, risk detection, and AI agents & Data MCPs. [Explore more use cases](https://datahippo.ai/index.md) or [bring us a challenge](https://datahippo.ai/contact.md). --- # Demo availability and booking guide | Data Hippo Source: https://datahippo.ai/docs/booking/ # Demo availability and booking guide Read available times with a simple API, then book a founder-led demo with the attendee’s approval. This public interface schedules a 30-minute Google Meet demonstration. It does not access customer data, the Data Hippo platform, or an administrative calendar. [Read this guide as Markdown →](https://datahippo.ai/docs/booking.md) · [OpenAPI specification →](https://datahippo.ai/openapi.json) ## When should an agent use this interface? Use [the platform guide](https://datahippo.ai/docs/index.md) and [product reference](https://datahippo.ai/docs/products.md) to answer questions about Data Hippo. Use `GET /demo/slots` only when someone wants to see meeting times. Use `POST /demo/book` only when the attendee has approved one specific time and their contact details. Reading the website is not permission to book. This is a public demo scheduler, not a developer API for querying healthcare data or controlling a customer deployment. If the request needs those capabilities, [contact the team](https://datahippo.ai/contact.md) rather than inventing an endpoint. ## Read availability, then book an approved time These examples use the production availability endpoint. Discover the current endpoint from `servers[0].url` in the [OpenAPI specification](https://datahippo.ai/openapi.json). A locally served specification points to development infrastructure. Both environments use the real calendar; development is not a booking dry run. Run one example in your terminal. Each reads the next seven days from the current UTC instant, using America/New\_York as the display timezone. Change that value to the attendee’s IANA timezone. No credential, package install, or forged browser headers are needed. By default, these examples only read availability; they never choose or book the first slot. After the attendee approves a returned time and their details, create `booking.json` using the template below. Replace every placeholder. Set `attendee_confirmed` to `true` only after approval. Optional fields are `organization` and `comments`. Never include health data or secrets. ```json { "start": "REPLACE_WITH_APPROVED_UTC_ISO_TIME", "timezone": "America/New_York", "name": "REPLACE_WITH_APPROVED_NAME", "email": "REPLACE_WITH_APPROVED_EMAIL", "attendee_confirmed": false } ``` Create and retain one `IDEMPOTENCY_KEY` for this exact booking: current Unix milliseconds, a period, and a UUIDv4. For example, JavaScript generates it with `Date.now() + '.' + require('node:crypto').randomUUID()`; Python uses `str(int(time.time() * 1000)) + '.' + str(uuid.uuid4())`; PowerShell uses `[string][DateTimeOffset]::UtcNow.ToUnixTimeMilliseconds() + '.' + [guid]::NewGuid().ToString()`. Bash can use the Node or Python command. Save the resulting value before submission; do not regenerate it for a retry. Set the environment variables `IDEMPOTENCY_KEY` to that value and `BOOKING_CONFIRMED=yes` only when ready to submit the approved request. Rerun the same example. It reads availability again, then sends your chosen payload once. Keep responses private: they may contain attendee cancellation and rescheduling links. Save JavaScript as `availability.cjs` and run `node availability.cjs`. Save Python as `availability.py` and run `python3 availability.py`. Paste the Bash or PowerShell example into its respective shell. ### JavaScript · Node.js ```javascript // Node.js: built-in https; no packages or browser headers. const https = require('node:https'); const url = new URL('https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/slots'); url.search = new URLSearchParams({ after: new Date().toISOString(), days: '7', timezone: 'America/New_York' }); const request = https.get(url, { timeout: 15000 }, response => { let body = ''; response.setEncoding('utf8'); response.on('data', chunk => { body += chunk; }); response.on('end', () => { if (response.statusCode !== 200) { console.error('HTTP', response.statusCode, body); if (response.statusCode === 429) console.error('Retry after:', response.headers['retry-after'] || '60', 'seconds'); process.exitCode = 1; return; } console.log(body); // Optional second step, only after the attendee approves booking.json. if (process.env.BOOKING_CONFIRMED !== 'yes') return; const key = process.env.IDEMPOTENCY_KEY; if (!key) throw new Error('Set and retain IDEMPOTENCY_KEY first'); const payload = require('node:fs').readFileSync('booking.json', 'utf8'); if (JSON.parse(payload).attendee_confirmed !== true) throw new Error('Approval required'); const booking = https.request(new URL('/v1/demo/book', url), { method: 'POST', timeout: 20000, headers: { 'Content-Type': 'application/json', 'Idempotency-Key': key } }, result => { let receipt = ''; result.setEncoding('utf8'); result.on('data', chunk => { receipt += chunk; }); result.on('end', () => { console.log('Booking HTTP', result.statusCode, receipt); if (result.statusCode !== 200) process.exitCode = 1; }); }); booking.on('timeout', () => booking.destroy(new Error('Booking status unknown; retain the same key and payload'))); booking.on('error', error => { console.error(error.message); process.exitCode = 1; }); booking.end(payload); }); }); request.on('timeout', () => request.destroy(new Error('Availability request timed out'))); request.on('error', error => { console.error(error.message); process.exitCode = 1; }); ``` ### Python · standard library ```python from datetime import datetime, timezone from urllib.parse import urlencode from urllib.request import Request, urlopen from urllib.error import HTTPError, URLError import sys import os import json params = urlencode({ "after": datetime.now(timezone.utc).isoformat(), "days": 7, "timezone": "America/New_York" }) url = "https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/slots?" + params try: request = Request(url, headers={"User-Agent": "DataHippoAvailabilityExample/1.0"}) with urlopen(request, timeout=15) as response: print(response.read().decode("utf-8")) # Optional second step, only after the attendee approves booking.json. if os.environ.get("BOOKING_CONFIRMED") == "yes": key = os.environ["IDEMPOTENCY_KEY"] # Retain this key for this exact request. with open("booking.json", encoding="utf-8") as source: payload = json.load(source) if payload.get("attendee_confirmed") is not True: raise ValueError("Approval required") booking = Request("https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/book", data=json.dumps(payload).encode(), method="POST", headers={"Content-Type": "application/json", "Idempotency-Key": key, "User-Agent": "DataHippoAvailabilityExample/1.0"}) with urlopen(booking, timeout=20) as response: print(response.read().decode("utf-8")) except HTTPError as error: print("HTTP", error.code, error.read().decode("utf-8"), file=sys.stderr) if error.code == 429: print("Retry after:", error.headers.get("Retry-After", "60"), "seconds", file=sys.stderr) sys.exit(1) except (URLError, TimeoutError, ValueError, KeyError, OSError) as error: print(str(error), file=sys.stderr) sys.exit(1) ``` ### Bash · curl and date ```bash # macOS or Linux. curl prints response headers (including Retry-After) and JSON. # --fail-with-body requires curl 7.76 or later; no jq or packages required. curl --silent --show-error --fail-with-body --include --max-time 15 \ --get 'https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/slots' \ --data-urlencode "after=$(date -u '+%Y-%m-%dT%H:%M:%SZ')" \ --data-urlencode 'days=7' \ --data-urlencode 'timezone=America/New_York' || exit $? # Optional second step. Review booking.json with the attendee before setting yes. if [ "${BOOKING_CONFIRMED:-}" = yes ]; then : "${IDEMPOTENCY_KEY:?Set and retain IDEMPOTENCY_KEY first}" curl --silent --show-error --fail-with-body --include --max-time 20 \ 'https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/book' \ -H 'Content-Type: application/json' -H "Idempotency-Key: $IDEMPOTENCY_KEY" \ --data-binary @booking.json fi ``` ### PowerShell · native cmdlet ```powershell # PowerShell 5.1 or 7+, no modules required. $after = [Uri]::EscapeDataString([DateTimeOffset]::UtcNow.ToString("o")) $zone = [Uri]::EscapeDataString("America/New_York") $url = "https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/slots?after=$after&days=7&timezone=$zone" try { $result = Invoke-RestMethod -Uri $url -Method Get -TimeoutSec 15 -ErrorAction Stop $result | ConvertTo-Json -Depth 5 # Optional second step, only after the attendee approves booking.json. if ($env:BOOKING_CONFIRMED -eq 'yes') { if (-not $env:IDEMPOTENCY_KEY) { throw 'Set and retain IDEMPOTENCY_KEY first' } $payload = Get-Content -Raw booking.json if (($payload | ConvertFrom-Json).attendee_confirmed -ne $true) { throw 'Approval required' } $booking = @{ Uri = 'https://dh-booking-prod.fritz-9f9.workers.dev/v1/demo/book' Method = 'Post'; ContentType = 'application/json'; Body = $payload Headers = @{ 'Idempotency-Key' = $env:IDEMPOTENCY_KEY }; TimeoutSec = 20; ErrorAction = 'Stop' } Invoke-RestMethod @booking | ConvertTo-Json -Depth 5 } } catch { $response = $_.Exception.Response if ($response) { Write-Error ("HTTP " + [int]$response.StatusCode + ": " + $_.ErrorDetails.Message) if ([int]$response.StatusCode -eq 429) { Write-Host "Wait at least 60 seconds before requesting availability again." } } else { Write-Error $_.Exception.Message } exit 1 } ``` The JSON response contains `slots` with UTC `iso` start times, and `earliest`/`latest` bounds in Unix milliseconds. An empty list means no availability in that interval. A read does not reserve a time. Calendly owns minimum notice, working hours, and calendar conflicts; do not infer a fixed notice period. Only complete months within the next 90 days are offered. Availability requests time out after 15 seconds; booking requests after 20 seconds. For HTTP 429, wait the number of seconds in `Retry-After` before another read; if absent, wait at least 60 seconds. For other failures, show the error instead of presenting stale slots as available. Do not repeatedly poll or automatically retry booking submissions. ## Book with an agent or in a browser 1. Read availability, then present a few choices in the attendee’s timezone. Recheck availability before final selection; times can change. 2. Get the attendee’s explicit approval of the date, time, timezone, name, and email before submitting. Never submit health data or secrets. 3. POST the approved request to `/demo/book` with `attendee_confirmed: true` and the retained `Idempotency-Key`, or use the browser calendar at `https://datahippo.ai/contact/`. Both create a real invitation. 4. HTTP 200 confirms the booking. The same key and unchanged payload replay the original result, not another invitation. After a timeout, retain the exact key and payload; never generate a replacement key automatically. 5. Keys expire 24 hours after their timestamp. HTTP 409 with `idempotency_conflict` means the payload changed; `booking_unconfirmed` means the outcome is pending or uncertain; `idempotency_expired` means the key expired. For an uncertain or expired outcome, check the invitation or contact us before any new attempt. 6. Keep the invitee’s cancellation and rescheduling links private. Share them only with the attendee; they are not administrative links. Submission creates a real invitation. HTTP 429 requires waiting for `Retry-After`; HTTP 503 means scheduling is unavailable. Do not automatically retry submissions. If booking cannot be confirmed, give the attendee `https://datahippo.ai/contact/` and the proposed times instead. [Contact information →](https://datahippo.ai/contact.md) ## Errors, rate limits, and API changes API errors follow the `ApiError` schema: an `error` message, machine-readable `code`, and safe next action in `resolution`. Handle HTTP status first, then a recognized code; do not depend on exact message wording. Unknown errors should be shown safely to the requester, without treating the booking as confirmed. - **400:** Check required fields, explicit approval, and the timestamped key. Unknown booking fields are rejected. - **403:** The request origin is not allowed. Server clients omit `Origin`, `Sec-Fetch-Site`, and `Sec-Fetch-Dest`; native fetch `Sec-Fetch-Mode` is accepted. Do not spoof browser headers. - **404:** Check the method and route against the OpenAPI specification. - **409:** A slot or idempotency conflict needs attention. Pending or uncertain bookings must be checked before another attempt. - **413 / 415:** Send JSON with `Content-Type: application/json`, within the 4,096-byte request limit. - **429:** Respect `Retry-After` in seconds; the current limit window is 60 seconds. Avoid polling loops. Keep the same key and payload if a booking is retried. - **503:** Scheduling is unavailable. Do not interpret this as proof that an earlier booking did not happen. The API publishes `RateLimit-Limit` and `RateLimit-Policy`: 30 availability reads or 3 booking attempts per 60-second window, per client IP and Cloudflare location. Enforcement is distributed, not an exact global counter. No remaining-quota or reset timestamp is claimed; use `Retry-After` on HTTP 429. The live [OpenAPI document](https://datahippo.ai/openapi.json) is the current contract. Its server URL includes `/v1`; append the documented operation path. Legacy unversioned routes remain supported. Changes within v1 remain backward-compatible and additive; a breaking change requires a new major-version URL. Handle unfamiliar optional fields and error codes safely. `info.version` identifies the document revision. There is no fixed deprecation deadline or API support SLA; contact us about compatibility needs.