{"id":2771,"date":"2026-08-23T09:49:22","date_gmt":"2026-08-23T09:49:22","guid":{"rendered":"https:\/\/www.deepwraite.de\/?page_id=2771"},"modified":"2026-09-04T14:08:04","modified_gmt":"2026-09-04T14:08:04","slug":"ai-in-action","status":"publish","type":"page","link":"https:\/\/www.deepwraite.de\/en\/ai-in-action\/","title":{"rendered":"AI in Action"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-page\" data-elementor-id=\"2771\" class=\"elementor elementor-2771\" data-elementor-post-type=\"page\">\n\t\t\t\t\t\t<section class=\"elementor-section elementor-top-section elementor-element elementor-element-08a30a1 elementor-section-boxed elementor-section-height-default elementor-section-height-default\" data-id=\"08a30a1\" data-element_type=\"section\" data-e-type=\"section\">\n\t\t\t\t\t\t<div class=\"elementor-container elementor-column-gap-default\">\n\t\t\t\t\t<div class=\"elementor-column elementor-col-100 elementor-top-column elementor-element elementor-element-764ccf0\" data-id=\"764ccf0\" data-element_type=\"column\" data-e-type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t<div class=\"elementor-element elementor-element-8a23eea elementor-widget elementor-widget-html\" data-id=\"8a23eea\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<style>\nbody{background:#111111}\n.page-header{display:none}\n.dwe{\n  background:#111111; overflow-x:clip;\n  --line:rgba(255,255,255,.08); --line-2:rgba(255,255,255,.16);\n  --ink:#ffffff; --ink-2:#b4b4b4; --ink-3:#7a8593;\n  --teal:#0ee2cf; --teal-2:#4fb6b1;\n  font-family:Inter,-apple-system,system-ui,\"Segoe UI\",Roboto,Arial,sans-serif;\n  color:var(--ink-2); font-size:17px; line-height:1.75;\n}\n.dwe *{box-sizing:border-box}\n.dwe h1{margin:0; line-height:1.15; font-size:clamp(38px,6vw,62px); font-weight:600; color:var(--teal); letter-spacing:-.005em}\n.dwe .wrap{width:100%; max-width:1160px; margin:0 auto; padding:0}\n@media(max-width:1200px){.dwe .wrap{padding:0 16px}}\n.dwe .hero{\n  position:relative; margin:0; padding:88px 0 76px; border-radius:18px;\n  background:\n    linear-gradient(180deg, rgba(17,17,17,.55) 0%, rgba(17,17,17,.82) 58%, #111 100%),\n    radial-gradient(900px 420px at 72% 18%, rgba(14,226,207,.16), transparent 65%),\n    url(\"https:\/\/www.deepwraite.de\/wp-content\/uploads\/2025\/07\/40-abstract-3d-shapes-WJL94QZ.png\") center\/cover no-repeat,\n    #161616;\n  overflow:hidden;\n}\n.dwe .hero::after{content:\"\"; position:absolute; left:0; right:0; bottom:0; height:1px;\n  background:linear-gradient(90deg, transparent, rgba(14,226,207,.4), transparent);}\n.dwe .lede{font-size:20px; line-height:1.55; color:#fff; margin:24px 0 0; max-width:62ch; font-weight:400}\n<\/style>\n\n<div class=\"dwe\">\n  <div class=\"hero\">\n    <div class=\"wrap\">\n      <h1>AI in Action<\/h1>\n      <p class=\"lede\">What actually runs at our clients &mdash; and at our own desk. Each system with the status it actually has &mdash; in operation, in development or prototype &mdash; together with the decisions behind them.<\/p>\n    <\/div>\n  <\/div>\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-46aa1f9 elementor-widget elementor-widget-html\" data-id=\"46aa1f9\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<style>.dw-panel__status{display:inline-block;margin-left:8px;padding:2px 10px;border-radius:999px;font-size:12px;font-weight:600;letter-spacing:.02em;line-height:1.6;vertical-align:baseline;border:1px solid;white-space:nowrap}.dw-panel__status--betrieb{color:#4fb6b1;border-color:rgba(79,182,177,.5);background:rgba(79,182,177,.12)}.dw-panel__status--aufbau{color:#e0b04f;border-color:rgba(224,176,79,.5);background:rgba(224,176,79,.12)}.dw-panel__status--proto{color:#d0d0d0;border-color:rgba(208,208,208,.45);background:rgba(208,208,208,.10)}.dw-panel__status--fertig{color:#9fb8d6;border-color:rgba(159,184,214,.45);background:rgba(159,184,214,.10)}<\/style><style>.dw-cases .dw-cases__inner{max-width:1160px;margin-left:auto;margin-right:auto}.dw-cases{padding-top:24px}<\/style><style>.dw-cases .dw-cases__eyebrow{font-size:30.4px;line-height:38px;font-weight:600}.dw-cases .dw-panel__title{font-size:19.2px;line-height:26.88px;font-weight:500}<\/style><section class=\"dw-cases\">\n  <div class=\"dw-cases__inner\">\n    <input class=\"dw-r\" type=\"radio\" name=\"dwcases\" id=\"dwc1\" checked>\n    <input class=\"dw-r\" type=\"radio\" name=\"dwcases\" id=\"dwc2\">\n    <input class=\"dw-r\" type=\"radio\" name=\"dwcases\" id=\"dwc3\">\n    <input class=\"dw-r\" type=\"radio\" name=\"dwcases\" id=\"dwc4\">\n    <input class=\"dw-r\" type=\"radio\" name=\"dwcases\" id=\"dwc5\">\n    <div class=\"dw-tabs\">\n      <label class=\"dw-tab\" for=\"dwc1\">Voice assistant<\/label>\n      <label class=\"dw-tab\" for=\"dwc2\">Chatbot<\/label>\n      <label class=\"dw-tab\" for=\"dwc3\">ProjektMelder<\/label>\n      <label class=\"dw-tab\" for=\"dwc4\">TenderFit<\/label>\n      <label class=\"dw-tab\" for=\"dwc5\">NO\u2082 forecasting<\/label>\n    <\/div>\n<article class=\"dw-panel\" id=\"dwp1\">\n      <label class=\"dw-panel__head\" for=\"dwc1\">\n        <span class=\"dw-panel__head-title\">Voice assistant<\/span>\n        <span class=\"dw-panel__head-icon\">+<\/span>\n      <\/label>\n      <div class=\"dw-panel__content\">\n        <p class=\"dw-panel__meta\">In-house build, Deepwraite\u00ae \u00b7 July\u2013August 2026 <span class=\"dw-panel__status dw-panel__status--proto\">Prototype<\/span><\/p>\n        <h3 class=\"dw-panel__title\">AI voice assistant with a cloned voice \u2014 disclosed from the first sentence<\/h3>\n        <div class=\"dw-panel__cols\">\n          <div class=\"dw-panel__body\">\n            <p>Our phone agent is designed to answer when we cannot: it greets the caller, understands the request, qualifies it and books an appointment on request \u2014 invitation included. It speaks with a cloned voice (<a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-voice-clone\">Voice Clone<\/a>), and that is precisely why the first sentence is no formality: <span class=\"dw-panel__kicker\">\u201eI sound like Kai, but I am an artificial intelligence.\"<\/span> The disclosure required by <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-art-50\">Article 50 EU AI Act<\/a> is not buried in our small print \u2014 it is the greeting. As things stand it is a prototype: the pipeline runs, but we will not switch on a phone number until governance and data protection are fully settled &mdash; exactly the order we recommend to our clients.<\/p>\n            <p>Built as our own real-time pipeline with response times under 800 milliseconds, bilingual in German and English, designed for end-to-end data processing inside the EU. It shares its knowledge base with our website chatbot: one source, two channels. And when there is no human at the other end but another <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-agent\">AI agent<\/a>, it switches via Gibberlink to a tone-based protocol \u2014 two machines that stop reciting pleasantries and simply exchange data.<\/p>\n            <p>It is allowed to do more than the chatbot: make contact and book appointments. What it is not allowed to do: commit to prices or interpret the law.<\/p>\n          <\/div>\n          <div>\n            <figure class=\"dw-figure\">\n              <span class=\"dw-schema dw-schema--1\"><\/span>\n              <figcaption>Anruf rein, Termin raus \u2014 in unter 800 Millisekunden Antwortzeit.<\/figcaption>\n            <\/figure>\n            <aside class=\"dw-facts\">\n              <p class=\"dw-facts__title\">At a glance<\/p>\n              <dl>\n                <div><dt>Role<\/dt><dd>Konzeption &amp; Entwicklung, eigenes Produkt<\/dd><\/div>\n                <div><dt>Kanal<\/dt><dd>Telefon, eingehend<\/dd><\/div>\n                <div><dt>Sprachen<\/dt><dd>Deutsch, Englisch, Gibberlink (Agent zu Agent)<\/dd><\/div>\n                <div><dt>Antwortzeit<\/dt><dd>unter 800 ms<\/dd><\/div>\n                <div><dt>Offenlegung<\/dt><dd>Article 50 EU AI Act, im ersten Satz<\/dd><\/div>\n              <\/dl>\n              <p class=\"dw-facts__tech\">TENIOS (SIP, DE) \u00b7 LiveKit Agents (Python) \u00b7 Deepgram (STT) \u00b7 OpenAI GPT-4o \u00b7 ElevenLabs (voice) \u00b7 Google Calendar API \u00b7 Gibberlink. An EU-sovereign build (Mistral, EU STT, self-hosted TTS) is prepared.<\/p>\n              <p class=\"dw-facts__more\"><a href=\"\/en\/governance-glossary\/\">Terms explained in the Governance Glossary \u2192<\/a><\/p>\n            <\/aside>\n          <\/div>\n        <\/div>\n      <\/div>\n    <\/article>\n<article class=\"dw-panel\" id=\"dwp2\">\n      <label class=\"dw-panel__head\" for=\"dwc2\">\n        <span class=\"dw-panel__head-title\">Chatbot<\/span>\n        <span class=\"dw-panel__head-icon\">+<\/span>\n      <\/label>\n      <div class=\"dw-panel__content\">\n        <p class=\"dw-panel__meta\">In-house build, Deepwraite\u00ae \u00b7 July 2026 <span class=\"dw-panel__status dw-panel__status--betrieb\">in operation<\/span><\/p>\n        <h3 class=\"dw-panel__title\">AI chatbot: data minimisation as a design decision<\/h3>\n        <div class=\"dw-panel__cols\">\n          <div class=\"dw-panel__body\">\n            <p>Most website bots want your email address first. <span class=\"dw-panel__kicker\">Ours does not.<\/span> It answers your questions on AI governance, <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-eu-ai-act\">EU AI Act<\/a> and <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-iso-42001\">ISO 42001<\/a> \u2014 and when things get specific it points to an appointment instead of collecting contact details. Data minimisation here is not an afterthought but a design decision.<\/p>\n            <p>The chatbot is labelled as AI, listed in our internal AI register and has a defined change process \u2014 we treat our own bot exactly as we advise our clients to treat theirs. Its knowledge base is the same one the phone agent draws on: change a statement in one place and it changes in both channels. <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-guardrail\">Guardrails<\/a> keep it on topic \u2014 no price commitments, no case-by-case legal advice, no invented deadlines.<\/p>\n            <p>A chatbot that says what it is and does not ask for what it does not need.<\/p>\n          <\/div>\n          <div>\n            <figure class=\"dw-figure\">\n              <span class=\"dw-schema dw-schema--2\"><\/span>\n              <figcaption>Antworten geben, ohne Kontaktdaten einzusammeln.<\/figcaption>\n            <\/figure>\n            <aside class=\"dw-facts\">\n              <p class=\"dw-facts__title\">At a glance<\/p>\n              <dl>\n                <div><dt>Role<\/dt><dd>Konzeption &amp; Entwicklung, eigenes Produkt<\/dd><\/div>\n                <div><dt>Kanal<\/dt><dd>Website<\/dd><\/div>\n                <div><dt>Themen<\/dt><dd>KI-Governance, EU AI Act, ISO 42001<\/dd><\/div>\n                <div><dt>Datenerhebung<\/dt><dd>keine Kontaktdaten, Verweis auf Termin<\/dd><\/div>\n                <div><dt>Governance<\/dt><dd>als KI gekennzeichnet, im KI-Register gef\u00fchrt<\/dd><\/div>\n              <\/dl>\n              <p class=\"dw-facts__tech\">Anthropic Claude \u00b7 knowledge base as a single source of truth, shared with the phone agent \u00b7 embedded via WordPress\/Elementor.<\/p>\n              <p class=\"dw-facts__more\"><a href=\"\/en\/governance-glossary\/\">Terms explained in the Governance Glossary \u2192<\/a><\/p>\n            <\/aside>\n          <\/div>\n        <\/div>\n      <\/div>\n    <\/article>\n<article class=\"dw-panel\" id=\"dwp3\">\n      <label class=\"dw-panel__head\" for=\"dwc3\">\n        <span class=\"dw-panel__head-title\">ProjektMelder<\/span>\n        <span class=\"dw-panel__head-icon\">+<\/span>\n      <\/label>\n      <div class=\"dw-panel__content\">\n        <p class=\"dw-panel__meta\">Client project, fire safety \u00b7 March 2026 <span class=\"dw-panel__status dw-panel__status--betrieb\">in operation<\/span><\/p>\n        <h3 class=\"dw-panel__title\">ProjektMelder \u2014 tender monitoring for a fire safety specialist<\/h3>\n        <div class=\"dw-panel__cols\">\n          <div class=\"dw-panel__body\">\n            <p>Screening public procurement portals by hand every day costs sales time and still leaves gaps. For a client in fire safety we built an <a class=\"dw-gloss\" href=\"\/en\/governance-glossary\/#dwg-agent\">agent<\/a> that watches newly published procedures on the major platforms, matches them against service profile, region, contract volume and contracting authority \u2014 and reports only the hits that genuinely matter commercially. <span class=\"dw-panel__kicker\">The smoke-detector principle: don't watch constantly, get alerted when something burns.<\/span><\/p>\n            <p>Every hit arrives with direct links into the tender, procurement and organisational data, so the assessment starts where the research would otherwise begin. Hours of screening become a few minutes, relevant procedures enter the evaluation in structured form \u2014 and the pitch decision rests on clean data.<\/p>\n          <\/div>\n          <div>\n            <figure class=\"dw-figure\">\n              <span class=\"dw-schema dw-schema--3\"><\/span>\n              <figcaption>Aus mehreren Stunden Sichtung werden wenige Minuten.<\/figcaption>\n            <\/figure>\n            <aside class=\"dw-facts\">\n              <p class=\"dw-facts__title\">At a glance<\/p>\n              <dl>\n                <div><dt>Sector<\/dt><dd>Fire safety, public sector clients<\/dd><\/div>\n                <div><dt>Role<\/dt><dd>Concept &amp; development of the agent<\/dd><\/div>\n                <div><dt>Sources<\/dt><dd>several public procurement platforms, monitored daily<\/dd><\/div>\n                <div><dt>Filter logic<\/dt><dd>scope of services, region, contract volume, contracting authority<\/dd><\/div>\n                <div><dt>Outcome<\/dt><dd>hours of screening \u2192 a few minutes, pre-sorted and with deep links<\/dd><\/div>\n              <\/dl>\n              <p class=\"dw-facts__more\"><a href=\"\/en\/governance-glossary\/\">Terms explained in the Governance Glossary \u2192<\/a><\/p>\n            <\/aside>\n          <\/div>\n        <\/div>\n      <\/div>\n    <\/article>\n<article class=\"dw-panel\" id=\"dwp4\">\n      <label class=\"dw-panel__head\" for=\"dwc4\">\n        <span class=\"dw-panel__head-title\">TenderFit<\/span>\n        <span class=\"dw-panel__head-icon\">+<\/span>\n      <\/label>\n      <div class=\"dw-panel__content\">\n        <p class=\"dw-panel__meta\">Client project, advertising \u00b7 February 2026 <span class=\"dw-panel__status dw-panel__status--betrieb\">in operation<\/span><\/p>\n        <h3 class=\"dw-panel__title\">TenderFit \u2014 go or no-go on complex tenders<\/h3>\n        <div class=\"dw-panel__cols\">\n          <div class=\"dw-panel__body\">\n            <p><span class=\"dw-panel__kicker\">Not every tender you could win is one you should pitch for.<\/span> TenderFit reads tender documents in a structured way \u2014 lots, budgets, eligibility criteria, risks \u2014 and scores the strategic fit through a weighted model of capability fit, references, capacity, commercial viability and risk.<\/p>\n            <p>The result is a reasoned go, maybe or no-go recommendation, including an estimate of whether the pitch effort pays off. For multi-lot procedures the model also simulates consortia with lead and partner logic across different trades. Developed for an agency \u2014 transferable to any sector where public tenders tie up a lot of preparatory work.<\/p>\n          <\/div>\n          <div>\n            <figure class=\"dw-figure\">\n              <span class=\"dw-schema dw-schema--4\"><\/span>\n              <figcaption>Weighted scoring instead of gut feeling \u2014 with a reasoned recommendation.<\/figcaption>\n            <\/figure>\n            <aside class=\"dw-facts\">\n              <p class=\"dw-facts__title\">At a glance<\/p>\n              <dl>\n                <div><dt>Sector<\/dt><dd>Advertising \/ creative agency<\/dd><\/div>\n                <div><dt>Format<\/dt><dd>specialised GPT<\/dd><\/div>\n                <div><dt>Analysis<\/dt><dd>lots, budgets, eligibility criteria, risks<\/dd><\/div>\n                <div><dt>Scoring<\/dt><dd>capability fit, references, capacity, commercial viability, risk<\/dd><\/div>\n                <div><dt>Outcome<\/dt><dd>go \/ maybe \/ no-go incl. consortium simulation<\/dd><\/div>\n              <\/dl>\n              <p class=\"dw-facts__more\"><a href=\"\/en\/governance-glossary\/\">Terms explained in the Governance Glossary \u2192<\/a><\/p>\n            <\/aside>\n          <\/div>\n        <\/div>\n      <\/div>\n    <\/article>\n<article class=\"dw-panel\" id=\"dwp5\">\n      <label class=\"dw-panel__head\" for=\"dwc5\">\n        <span class=\"dw-panel__head-title\">NO\u2082 forecasting<\/span>\n        <span class=\"dw-panel__head-icon\">+<\/span>\n      <\/label>\n      <div class=\"dw-panel__content\">\n        <p class=\"dw-panel__meta\">Machine learning project \u00b7 January 2026 <span class=\"dw-panel__status dw-panel__status--fertig\">completed<\/span><\/p>\n        <h3 class=\"dw-panel__title\">NO\u2082 forecasting with gradient boosting<\/h3>\n        <div class=\"dw-panel__cols\">\n          <div class=\"dw-panel__body\">\n            <p><span class=\"dw-panel__kicker\">Not every task needs a language model.<\/span> To forecast NO\u2082 concentrations we trained a classic machine learning model on historical environmental and measurement data: data preparation and feature engineering, gradient boosting, validation and performance tuning, with a persisted model for production use.<\/p>\n            <p>The result is a reliable forecast of air pollutant levels as a basis for data-driven decisions in environmental and mobility contexts. We choose the method to fit the question \u2014 not to fit the headlines.<\/p>\n          <\/div>\n          <div>\n            <figure class=\"dw-figure\">\n              <span class=\"dw-schema dw-schema--5\"><\/span>\n              <figcaption>Historical measurement data turned into a reliable forecast.<\/figcaption>\n            <\/figure>\n            <aside class=\"dw-facts\">\n              <p class=\"dw-facts__title\">At a glance<\/p>\n              <dl>\n                <div><dt>Method<\/dt><dd>gradient boosting (supervised learning)<\/dd><\/div>\n                <div><dt>Data<\/dt><dd>historical environmental and measurement data<\/dd><\/div>\n                <div><dt>Scope<\/dt><dd>data preparation, feature engineering, training, validation<\/dd><\/div>\n                <div><dt>Artefact<\/dt><dd>persisted model (.joblib) for continued operation<\/dd><\/div>\n                <div><dt>Field of use<\/dt><dd>Environment and mobility<\/dd><\/div>\n              <\/dl>\n              <p class=\"dw-facts__more\"><a href=\"\/en\/governance-glossary\/\">Terms explained in the Governance Glossary \u2192<\/a><\/p>\n            <\/aside>\n          <\/div>\n        <\/div>\n      <\/div>\n    <\/article>\n  <\/div>\n<\/section>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d7efd48 elementor-widget elementor-widget-html\" data-id=\"d7efd48\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<style>.dwe .dwbtn,.dwe .dwbtn:link,.dwe .dwbtn:visited{display:inline-block;background:#4fb6b1;color:#111111;border:2px solid #4fb6b1;border-radius:50px;padding:15px 30px;font-size:16px;font-weight:500;line-height:1.2;text-decoration:none;transition:.3s}.dwe .dwbtn:hover{background:transparent;color:#4fb6b1;text-decoration:none}<\/style><div class=\"dwe\">\n  <div class=\"wrap\">\n    <div style=\"position:relative;margin:76px 0 0;padding:64px 0 40px;border-top:1px solid rgba(255,255,255,.08);background:radial-gradient(700px 260px at 20% 0%, rgba(14,226,207,.10), transparent 70%)\">\n      <h2 style=\"margin:0 0 14px;line-height:1.15;font-size:30.4px;font-weight:600;color:#fff\">Does this sound like your situation?<\/h2>\n      <p style=\"margin:0 0 28px;max-width:58ch;font-size:17px;line-height:1.75;color:#b4b4b4\">Every one of these projects started with an open question, not a finished specification. Tell us yours &mdash; a first conversation takes 30 to 45 minutes and commits you to nothing.<\/p>\n      <a class=\"dwbtn\" href=\"\/en\/contact\/\">Get in touch<\/a>\n    <\/div>\n  <\/div>\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>AI in Action What actually runs at our clients &mdash; and at our own desk. Each system with the status it actually has &mdash; in operation, in development or prototype &mdash; together with the decisions behind them. Voice assistant Chatbot ProjektMelder TenderFit NO\u2082 forecasting Voice assistant + In-house build, Deepwraite\u00ae \u00b7 July\u2013August 2026 Prototype AI [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"elementor_header_footer","meta":{"footnotes":""},"class_list":["post-2771","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/pages\/2771","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/comments?post=2771"}],"version-history":[{"count":10,"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/pages\/2771\/revisions"}],"predecessor-version":[{"id":3220,"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/pages\/2771\/revisions\/3220"}],"wp:attachment":[{"href":"https:\/\/www.deepwraite.de\/en\/wp-json\/wp\/v2\/media?parent=2771"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}