{"id":59096,"date":"2026-10-14T10:38:59","date_gmt":"2026-10-14T00:38:59","guid":{"rendered":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/14\/moving-ai-agents-into-production-safely-with-microsoft-foundry\/"},"modified":"2026-10-14T10:40:11","modified_gmt":"2026-10-14T00:40:11","slug":"moving-ai-agents-into-production-safely-with-microsoft-foundry","status":"publish","type":"post","link":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/14\/moving-ai-agents-into-production-safely-with-microsoft-foundry\/","title":{"rendered":"Moving AI Agents into Production Safely with Microsoft Foundry"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">In this blog post Moving AI Agents into Production Safely with Microsoft Foundry we will explain how Australian mid-market organisations can turn a promising AI agent pilot into a secure, controlled and measurable production service.<\/p>\n\n\n\n<!--more-->\n\n\n\n<p class=\"wp-block-paragraph\">The difficult part is rarely getting an agent to work in a demonstration. The difficult part is deciding what it can access, what it can do, who is accountable when it makes a mistake and how the business prevents one successful pilot from becoming dozens of unmanaged AI tools.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Microsoft Foundry provides a common platform for building, testing, deploying and governing AI applications and agents. An AI agent is software that can interpret a request, use approved information and tools, and complete tasks on a user&#8217;s behalf. Foundry brings together the AI models, business connections, security controls, evaluations and monitoring needed to manage that process.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Why successful AI pilots often stall<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">A pilot normally involves a small group, carefully selected data and close supervision. Production is different. Hundreds of employees may use the agent, sensitive information may pass through it and mistakes can affect customers, finances or operations.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This creates questions that were easy to avoid during the pilot:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>Can the agent read every document the user can access?<\/li><li>Can it update a customer record or only suggest an update?<\/li><li>Which AI models are approved?<\/li><li>Who reviews incorrect or unsafe responses?<\/li><li>How much can the agent spend each month?<\/li><li>Can the organisation prove what happened during an audit?<\/li><\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">These are governance questions, not simply software questions. The answer is to establish clear boundaries before expanding access.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If your organisation is still defining its overall platform, our article on designing an enterprise AI platform with Microsoft Foundry covers the broader architecture. Here, we will concentrate on the boundaries that make production agents manageable.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What a governance boundary means in practice<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">A governance boundary is a rule that limits where an agent operates and what it is allowed to do. Think of it as combining a job description, an access pass, a spending limit and an audit trail.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Microsoft Foundry can separate workloads into resources and projects. At a high level, the resource provides a shared management boundary, while projects allow teams or use cases to be separated with their own access, connections, monitoring and costs.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For example, a human resources agent should not sit in the same loosely controlled project as an experimental marketing assistant. The HR agent handles sensitive employee information and therefore needs stricter access, testing, logging and approval requirements.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Seven boundaries to set before production<\/h2>\n\n\n\n<h3 class=\"wp-block-heading\">1. Start with a business boundary<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Give the agent one defined business outcome. \u201cHelp the service desk resolve approved Microsoft 365 requests\u201d is measurable. \u201cHelp everyone work faster\u201d is not.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Document the users, tasks, expected benefits and tasks that are specifically excluded. Assign a business owner who is accountable for results, rather than leaving ownership entirely with IT.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This prevents scope creep and makes value easier to measure. You can compare resolution time, employee effort, error rates and cost before and after deployment.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">2. Give the agent its own identity<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">An agent should not use a developer&#8217;s account or a shared password to reach business systems. Microsoft Foundry works with Microsoft Entra ID, Microsoft&#8217;s identity and access service, to give agents controlled identities and permissions.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Use role-based access control, which means giving people and agents only the access required for their role. A service desk agent might read device details from Microsoft Intune, which manages and secures company devices, without receiving permission to change security policies.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Foundry also supports specialised agent identities so agents can authenticate to approved tools without storing passwords in prompts or code. This improves visibility and makes access easier to remove when an agent is retired.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">3. Separate data access from action permissions<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Reading information and changing information are different risk levels. An agent may be allowed to summarise an account without being allowed to change payment details, approve a refund or delete a record.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Start with read-only access wherever possible. Add actions gradually, with transaction limits and approval steps for higher-risk work.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Our guide to connecting Foundry agents to business systems explains the integration side in more detail. The governance principle is simple: the agent should reach only the specific systems, records and functions needed for its approved purpose.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">4. Control where information can travel<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Production agents may process customer records, contracts, internal emails or employee information. Australian organisations need to understand what information is collected, why it is used, where it is processed and who can see the results.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Microsoft Foundry supports private networking, which keeps approved traffic within controlled cloud networks rather than exposing services directly to the public internet. A deny-by-default design can block unapproved outbound connections while still allowing access to authorised data sources.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Privacy also needs to be addressed before launch. The Office of the Australian Information Commissioner recommends particular care when personal or sensitive information is entered into AI systems. A privacy impact assessment is sensible when an agent handles customer, health, financial or employee data.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">5. Approve models and apply safety controls<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Not every model belongs in every workload. Some models may be suitable for experimentation but inappropriate for a production process involving sensitive data or important decisions.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Foundry can use Azure Policy, Microsoft&#8217;s rules engine for cloud environments, to restrict deployments to approved models and publishers. Policies can also block models based on factors such as lifecycle status, helping prevent teams from quietly placing preview technology into critical production services.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Safety controls should inspect user requests and agent responses for risks such as harmful content, attacks designed to manipulate the agent or accidental disclosure. These controls reduce risk, but they do not replace testing or human judgement.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">6. Require evidence before release<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">A convincing demonstration is not a production test. Build a test set covering normal requests, incorrect information, ambiguous instructions, restricted data and deliberate attempts to bypass controls.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Foundry evaluations can measure response quality, whether answers are supported by approved information, task completion and safety. Set minimum release thresholds and keep the results as evidence for the production approval.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For high-impact actions, retain a person in the approval process. The agent can prepare a refund, contract response or account change, but an authorised employee approves it before execution.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If your technical team is implementing these controls in software, see building governed agents with Microsoft Agent Framework.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">7. Monitor quality, security and cost after launch<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Production approval is not the end of governance. Agent behaviour can change as business data, user behaviour, connected tools and AI models change.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Monitor failed tasks, slow responses, unusual access attempts, user complaints and the number of requests that require human help. Foundry observability, meaning the ability to see how the agent is behaving, can track operational performance, token consumption, errors and quality measures.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Set budgets and usage alerts as well. A useful agent can attract far more demand than expected, while a faulty workflow can repeatedly call an expensive model or tool. Cost boundaries prevent a productivity project from becoming an unexplained cloud bill.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">A representative mid-market scenario<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Consider an Australian services company with 180 employees. Its pilot agent successfully answers policy questions and helps staff prepare customer correspondence, but the pilot uses broad document access and has no formal spending limit.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Instead of releasing it company-wide, the organisation creates separate Foundry projects for internal policy assistance and customer operations. The policy agent receives read-only access to approved SharePoint libraries. The customer agent can retrieve account details, but proposed changes require employee approval.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The company also restricts approved models, adds private network connections, tests common failure scenarios and monitors cost per completed task. The outcome is not simply a safer agent. Management can see whether it is reducing search time, shortening response times and delivering enough value to justify expansion.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">How this supports Australian security expectations<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The Essential Eight, the Australian government&#8217;s baseline cybersecurity framework, was not designed specifically for AI. However, its focus on access control, system hardening, patching, backups and limiting administrative privileges remains highly relevant to the systems around an agent.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">An AI agent should not become a shortcut around existing security controls. Its identity, connected applications, supporting infrastructure and administrator access should be managed with the same discipline as any other important business system.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">A practical production approval checklist<\/h2>\n\n\n\n<ol class=\"wp-block-list\"><li>Define one measurable business outcome and assign an owner.<\/li><li>Separate the workload into an appropriate Foundry project.<\/li><li>Give users and agents the minimum required access.<\/li><li>Document approved data sources and prohibited information.<\/li><li>Restrict model deployment to an approved list.<\/li><li>Require human approval for financial, legal or high-impact actions.<\/li><li>Test quality, safety, privacy and failure scenarios.<\/li><li>Enable operational, security and cost monitoring.<\/li><li>Create an incident response and agent shutdown process.<\/li><li>Review permissions, performance and business value regularly.<\/li><\/ol>\n\n\n\n<h2 class=\"wp-block-heading\">Move forward without losing control<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The organisations getting value from AI agents are not removing governance to move faster. They are creating clear boundaries so teams can build confidently without renegotiating security, privacy and cost controls for every project.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That approach builds on the operating model discussed in moving AI agents from proof of concept to production. The goal is a repeatable path where experimentation remains easy, but production access must be earned through evidence.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">CloudProInc brings more than 20 years of enterprise IT experience to this work as a Microsoft Partner and Wiz Security Integrator. Our Melbourne-based team works hands-on across Microsoft Foundry, Azure, Microsoft 365, Defender and Wiz to connect AI governance with the security controls organisations already operate.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If your AI agent pilot works but the path to production is still unclear, we are happy to review the architecture, risks and governance boundaries with you \u2014 no strings attached.<\/p>\n\n\n","protected":false},"excerpt":{"rendered":"<p>A practical governance model for moving Microsoft Foundry agents beyond pilot while controlling data access, business actions, security, cost and accountability.<\/p>\n","protected":false},"author":1,"featured_media":59098,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":true,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_post_was_ever_published":false},"categories":[80,121,13,115],"tags":[],"class_list":["post-59096","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-agents","category-ai-governance-risk-management","category-blog","category-microsoft-ai-foundry"],"jetpack-related-posts":[{"id":58060,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale\/","url_meta":{"origin":59096,"position":0},"title":"How to Deploy LangGraph Agents to Microsoft Foundry Safely at Scale","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"Learn how to move LangGraph and other AI agents into Microsoft Foundry while controlling security, operating costs, system access and production risk.","rel":"","context":"In &quot;AI Agents&quot;","block_context":{"text":"AI Agents","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/ai-agents\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png 1x, \/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png 1.5x, \/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png 2x, \/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png 3x, \/wp-content\/uploads\/2026\/07\/how-to-deploy-langgraph-agents-to-microsoft-foundry-safely-at-scale.png 4x"},"classes":[]},{"id":58635,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a\/","url_meta":{"origin":59096,"position":1},"title":"From AI Proof of Concept to Production Using Foundry and A2A","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"A practical operating model for moving Azure AI agents from promising demos into secure, measurable production services using Microsoft Foundry and the A2A protocol.","rel":"","context":"In &quot;AI Agents&quot;","block_context":{"text":"AI Agents","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/ai-agents\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png 1x, \/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png 1.5x, \/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png 2x, \/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png 3x, \/wp-content\/uploads\/2026\/08\/from-ai-proof-of-concept-to-production-using-foundry-and-a2a.png 4x"},"classes":[]},{"id":57809,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely\/","url_meta":{"origin":59096,"position":2},"title":"Deploy Containerised Hosted Agents with Microsoft Foundry Safely","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"A practical guide for tech leaders on when containerised hosted agents make sense, how they work, and how to deploy them without creating security or cost surprises.","rel":"","context":"In &quot;Blog&quot;","block_context":{"text":"Blog","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/blog\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png 1x, \/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png 1.5x, \/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png 2x, \/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png 3x, \/wp-content\/uploads\/2026\/07\/deploy-containerised-hosted-agents-with-microsoft-foundry-safely.png 4x"},"classes":[]},{"id":57836,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service\/","url_meta":{"origin":59096,"position":3},"title":"Implementing Secure A2A Auth in Microsoft Foundry Agent Service","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"AI agents are starting to call other agents. Here is how to implement A2A authentication in Microsoft Foundry Agent Service without creating a security blind spot.","rel":"","context":"In &quot;Blog&quot;","block_context":{"text":"Blog","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/blog\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png 1x, \/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png 1.5x, \/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png 2x, \/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png 3x, \/wp-content\/uploads\/2026\/07\/implementing-secure-a2a-auth-in-microsoft-foundry-agent-service.png 4x"},"classes":[]},{"id":57787,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints\/","url_meta":{"origin":59096,"position":4},"title":"Connecting Microsoft Foundry Agents to External A2A Endpoints","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"A practical guide for tech leaders on connecting Microsoft Foundry agents to external A2A endpoints safely, without creating cost, security, or governance surprises.","rel":"","context":"In &quot;Blog&quot;","block_context":{"text":"Blog","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/blog\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png 1x, \/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png 1.5x, \/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png 2x, \/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png 3x, \/wp-content\/uploads\/2026\/07\/connecting-microsoft-foundry-agents-to-external-a2a-endpoints.png 4x"},"classes":[]},{"id":58020,"url":"https:\/\/www.cloudproinc.com.au\/index.php\/2026\/10\/12\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code\/","url_meta":{"origin":59096,"position":5},"title":"Build AI Agents Faster with Microsoft Foundry Toolkit in VS Code","author":"CPI Staff","date":"October 12, 2026","format":false,"excerpt":"Microsoft Foundry Toolkit helps development teams build and test business AI agents faster. Learn how to reduce delivery time without losing control of security, quality or costs.","rel":"","context":"In &quot;AI Agents&quot;","block_context":{"text":"AI Agents","link":"https:\/\/www.cloudproinc.com.au\/index.php\/category\/ai-agents\/"},"img":{"alt_text":"","src":"\/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png","width":350,"height":200,"srcset":"\/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png 1x, \/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png 1.5x, \/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png 2x, \/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png 3x, \/wp-content\/uploads\/2026\/07\/build-ai-agents-faster-with-microsoft-foundry-toolkit-in-vs-code.png 4x"},"classes":[]}],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"\/wp-content\/uploads\/2026\/10\/moving-ai-agents-into-production-safely-with-microsoft-foundry.png","_links":{"self":[{"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/posts\/59096","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/comments?post=59096"}],"version-history":[{"count":1,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/posts\/59096\/revisions"}],"predecessor-version":[{"id":59097,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/posts\/59096\/revisions\/59097"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/media\/59098"}],"wp:attachment":[{"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/media?parent=59096"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/categories?post=59096"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.cloudproinc.com.au\/index.php\/wp-json\/wp\/v2\/tags?post=59096"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}