# AI Employee Foundry: Full Public Knowledge Index Provider: Cloud Radix Canonical site: https://aiemployeefoundry.com Last verified: 2026-10-04 Maturity vocabulary: available, managed_pilot, roadmap, research. Maturity labels are part of the claim and must not be removed when summarizing. ## From role specification to governed operation Canonical: https://aiemployeefoundry.com/how-it-works AI Employee Foundry turns a business role into a governed operating system through discovery, authority design, controlled connections, acceptance testing, managed deployment, and measured improvement. ### Role before tools We define mission, owner, reporting line, voice, scorecard, systems, prohibitions, and escalation paths before choosing automation. ### Acceptance before operation The employee proves identity, permissions, workflows, failure behavior, and approval boundaries before it earns production responsibility. ### Work with evidence The operating model is designed to leave proof: what ran, what changed, what was approved, what failed, and what comes next. --- ## Choose a role. Then make it yours. Canonical: https://aiemployeefoundry.com/roles A Foundry role blueprint is a starting specification for recurring work, authority, systems, reporting, and outcomes. Cloud Radix adapts each blueprint to the company rather than pretending every business works the same way. ### Operations coordinator [available] Turns scattered requests, schedules, records, and follow-ups into an accountable operating rhythm. ### Marketing operator [available] Researches, drafts, publishes, measures, and improves useful content under editorial controls. ### Intake specialist [available] Captures, qualifies, routes, and documents inbound opportunities with clear escalation rules. ### Industry operator [available] Works inside a vertical operating system—legal, auctions, professional services, or another mapped environment. ### Managed node blueprint [managed_pilot] A repeatable private runtime pattern being proven through supervised customer deployments. ### Self-service activation [roadmap] Customer-led activation follows security, support, tenant-isolation, and synthetic-journey gates. --- ## Cloud, private, or hybrid—control follows the work Canonical: https://aiemployeefoundry.com/deployments AI Employee Foundry supports managed cloud, private/on-prem, and hybrid deployments. The right model depends on the systems the employee must reach, the data it handles, and the controls the business requires. ### Managed cloud [available] Best when the work is primarily cloud-based and the business values rapid managed operations. ### Managed private [available] Best when local access, physical custody, or a dedicated business-controlled node matters. ### Hybrid [available] Best when central management must coexist with approved local access and business-owned systems. --- ## A managed AI Employee without a new server room Canonical: https://aiemployeefoundry.com/deployments/cloud A managed cloud AI Employee runs on infrastructure operated for the customer by Cloud Radix, with scoped credentials, versioned releases, monitored work, and documented human authority. ### Good fit Cloud-first systems, distributed teams, managed maintenance, and workflows that do not require local network access. ### Control model Least-privilege connections, approval gates, release history, monitoring, and support remain part of the deployment. ### Not magic hosting Cloud deployment does not erase data, contractual, or workflow risk. Those boundaries are designed before launch. --- ## A dedicated AI Employee node under business control Canonical: https://aiemployeefoundry.com/deployments/private A managed private deployment places a dedicated runtime in a customer-approved environment while Cloud Radix manages the role, release, support, and governance model under agreed access controls. ### Dedicated environment The employee operates on a customer-specific node rather than a shared public application experience. ### Local reach Private deployment can support approved access to local files, devices, and business systems that cloud-only software cannot safely reach. ### Managed responsibility Physical location alone is not a security program. Updates, credentials, network boundaries, backup, and recovery still require disciplined operation. --- ## Central management. Approved local reach. Canonical: https://aiemployeefoundry.com/deployments/hybrid A hybrid AI Employee combines a managed control and release layer with a dedicated local runtime for approved private-system access. Authority remains explicit on both sides of the boundary. ### Bounded bridge Only required systems and actions cross the local/cloud boundary; the architecture is not an unrestricted tunnel. ### Unified role Mission, policies, release identity, and evidence stay coherent even when execution spans environments. ### Recovery-aware [roadmap] The managed-runtime roadmap includes replaceable infrastructure and verified restoration without permitting dual-active authority. --- ## Put the rules where the work happens Canonical: https://aiemployeefoundry.com/governance AI Employee governance defines who owns the role, what the employee may do, what requires approval, what is prohibited, how work is evidenced, and how authority can be paused or revoked. ### Authority matrix Read, draft, recommend, execute, publish, send, spend, and delete are separate powers—not one blanket permission. ### Approval gates Consequential work pauses for the right human, with enough context to approve, reject, or revise it. ### Prohibited actions The role specification records actions the employee must never take, even if technically possible. ### Evidence and audit Proof logs make operating decisions reviewable instead of burying them inside chat transcripts. ### Emergency control Credentials, schedules, connectors, and the runtime can be paused without improvising during an incident. ### Budget boundaries Usage, vendor spend, and external actions belong inside explicit operating limits. --- ## Connect only what the role needs Canonical: https://aiemployeefoundry.com/connectors Foundry connectors give an AI Employee scoped access to approved business systems. Each connection is selected for a defined workflow and labeled by maturity; no universal integration claim is made. ### Google Workspace [available] Gmail, Calendar, Drive, Docs, Sheets, Tasks, and related tools based on authorized scope and workflow. ### Web and CMS [available] Structured content, publishing, forms, analytics, search, and source-backed web operations. ### CRM and case systems [available] Customer, lead, matter, project, and activity workflows through approved APIs or custom integration. ### Communications [managed_pilot] Email, messaging, and phone workflows with identity, disclosure, approval, and escalation controls. ### Accounting and payments [managed_pilot] Read, reconciliation, drafting, and controlled actions require system-specific review and financial authority design. ### Connector marketplace [roadmap] A public self-service marketplace is not part of the current offer. --- ## Know what your AI Employee is doing Canonical: https://aiemployeefoundry.com/command-center Command Center is the operational layer Cloud Radix is assembling to manage AI Employee identity, work, health, policies, releases, costs, incidents, support, backup, and measurable business outcomes. ### Work and proof [managed_pilot] See recurring work, outcomes, approvals, exceptions, and proof-of-work rather than counting chat messages. ### Health and connectors [managed_pilot] Surface runtime state, credential health, integration failures, queues, and human attention needs. ### Release control [managed_pilot] Track what version is deployed, what changed, what passed acceptance, and what can be rolled back. ### Fleet self-service [roadmap] Customer-led multi-node activation follows production security and support gates. --- ## Design for failure before failure chooses the time Canonical: https://aiemployeefoundry.com/reliability Foundry reliability is based on controlled releases, rollback, isolated state, verified backups, restoration testing, and explicit recovery authority—not on unverified promises of zero downtime or zero data loss. ### Versioned releases [managed_pilot] Identity, configuration, skills, policies, and code move through an identified release rather than ad hoc machine edits. ### Acceptance evidence [managed_pilot] A release earns production responsibility through preflight and role-specific acceptance checks. ### Restore drills [roadmap] Recovery claims will become measurable only after repeated, evidenced restoration drills. ### Active/passive recovery [roadmap] The design prevents two copies from holding the same active authority during failover. --- ## Least privilege is a product decision Canonical: https://aiemployeefoundry.com/security Foundry security begins with role-level least privilege: isolated customer state, scoped credentials, separated public and operational systems, controlled releases, limited support access, and revocable authority. ### Tenant boundaries Customer identity, memory, state, secrets, evidence, and connectors are designed as customer-specific assets. ### Scoped credentials The employee receives only the accounts, permissions, and duration needed for its approved job. ### Release provenance Production changes should be traceable to a version, source, checks, and approval path. ### Public/private separation This marketing site cannot access operational tenant, fleet, backup, incident, or secret data. --- ## Start with the business, not the model Canonical: https://aiemployeefoundry.com/industries An industry AI Employee works best when it is embedded in the records, decisions, approvals, language, and operating rhythm of a specific business—not when it is deployed as a generic assistant. ### Legal [available] Personal-injury intake, case operations, records, documents, deadlines, communications, and source-backed content. ### Auctions [available] Seller intake through marketing, bidding, awards, invoicing, settlement, and the next sale. ### Architectural signage [managed_pilot] Lead and bid intake, research, content, estimating workflow, and operating-platform discovery. ### Professional services [available] Knowledge, intake, delivery, follow-up, reporting, and client experience across connected systems. --- ## An AI Employee inside the work of a law firm Canonical: https://aiemployeefoundry.com/industries/legal In a law-firm operating system, an AI Employee can support intake, records, tasks, documents, analysis, communication, deadlines, and publishing while attorneys retain legal judgment and client authority. ### James at Delventhal James works across the public website and the private case operating system as one governed role—not as a disconnected chatbot. ### Matter flow The operating scope spans potential clients, cases, contacts, providers, medical records, damages, demands, discovery, settlement, and follow-up. ### Attorney authority Legal advice, strategy, settlement positions, filings, and consequential communications remain under attorney review. --- ## One platform from seller intake to settlement Canonical: https://aiemployeefoundry.com/industries/auctions The Gavel Platform is an all-in-one auction software and marketing operating system that connects the auction house's back office, branded public marketplace, bidding engine, financial workflow, and Gavel AI Employee. ### Back office Contacts, sellers, projects, assets, marketing, bidder activity, awards, reports, invoices, settlement, and operating records. ### Public auction front end A branded bidder experience for discovery, registration, multi-parcel and timed-lot auctions, maximum proxy bids, soft closes, and results. ### Gavel AI Employee Gavel helps operate across the system so information becomes action instead of remaining trapped in software screens. --- ## Walter is learning the work before automating it Canonical: https://aiemployeefoundry.com/industries/architectural-signage Walter is a managed AI Employee pilot for Diskey Architectural Signage. The work begins with business discovery, governed marketing, lead and bid intake, and operating-system design—not with claims that unfinished estimating automation is already live. ### Discovery first [managed_pilot] Walter learns how opportunities arrive, how drawings and sign schedules are reviewed, who decides, and where evidence belongs. ### Growth operations [managed_pilot] The pilot includes useful SEO/AEO content, attributable lead paths, follow-up, research, and evidence-backed opportunity recommendations. ### Human bid authority People retain final bid/no-bid, scope, pricing, and proposal authority. --- ## Turn institutional knowledge into an operating advantage Canonical: https://aiemployeefoundry.com/industries/professional-services A professional-services AI Employee can organize intake, knowledge, research, delivery support, client follow-up, publishing, and reporting while professionals keep authority over advice and judgment. ### Intake and qualification Capture context once, route it correctly, and keep the next action visible. ### Knowledge in motion Turn approved expertise, documents, meetings, and prior work into useful drafts and operating support. ### Delivery rhythm Coordinate recurring tasks, follow-ups, reports, and proof without replacing the professional relationship. --- ## AI Employees built inside real businesses Canonical: https://aiemployeefoundry.com/case-studies Cloud Radix proves the AI Employee model by building named employees inside real operating environments. Each case study distinguishes current capability from pilot and roadmap work. ### Skywalker [available] Cloud Radix's first AI Employee helps grow and run the company and helps build the AI Employees that follow. ### James [available] A personal-injury AI Employee embedded across Delventhal Law Office's website and case operating system. ### Gavel [available] An auction AI Employee inside an all-in-one auction software, bidding, marketing, and settlement platform. --- ## The AI Employee helping build the company Canonical: https://aiemployeefoundry.com/case-studies/skywalker Skywalker is Cloud Radix's first AI Employee: a named digital operator who supports research, websites, software, marketing, operations, governance, deployments, documentation, and the creation of other AI Employees. ### Operates with the founders Skywalker works alongside Ken to turn conversations into plans, code, deployments, research, content, monitoring, and finished business assets. ### Builds the workforce Skywalker helped develop and refine Pistol Shrimp AI and continues helping establish James, Gavel, Walter, and future AI Employees. ### A voice in the room The spoken wake word “Skywalker” now begins a voice conversation—an early real-world glimpse of the kind of working relationship fiction imagined with JARVIS. --- ## James: an AI Employee inside a personal-injury operating system Canonical: https://aiemployeefoundry.com/case-studies/james James supports Delventhal Law Office across intake, case operations, records, documents, analysis, deadlines, communications, and publishing while legal judgment remains with the attorneys. ### One role, two surfaces The public website creates useful, source-backed client information. The private case system organizes the work that follows an inquiry. ### Deep operating scope The system covers potential clients, cases, contacts, providers, medical and accident analysis, damages, demands, discovery, expenses, settlement, deadlines, and more. ### Next stage [managed_pilot] Additional communications workflows require separate scoping, disclosure and human escalation. Phone answering is not included in a standard cloud role. --- ## Gavel: the AI Employee inside an auction operating system Canonical: https://aiemployeefoundry.com/case-studies/gavel Gavel connects an auction company's internal operations with its branded public bidding experience, so seller intake, cataloging, marketing, bidding, awards, invoices, settlement, and follow-up become one operating system. ### Built with operators The system was invented alongside Metzger Property Services' auction-industry expertise rather than designed from generic software assumptions. ### Combination-aware proxy bidding Gavel supports maximum proxy logic within multi-parcel bidding—a meaningful differentiator. Cloud Radix is not claiming “world's first” without broader independent verification. ### The employee inside the platform Gavel is not a chatbot placed beside auction software. The role is designed to work within the records, events, decisions, and next actions of the auction company. --- ## Build the operating system before opening the factory doors Canonical: https://aiemployeefoundry.com/roadmap AI Employee Foundry is moving from proven founder-led delivery to repeatable managed deployments, then toward customer activation only after security, isolation, support, acceptance, recovery, and economics are proven. ### Available now [available] Role discovery, custom implementation, cloud/private/hybrid deployment choices, authority design, and named AI Employees inside business operations. ### Managed pilot [managed_pilot] Standardized managed nodes, immutable releases, centralized health, support views, connector checks, and supervised operations. ### Roadmap [roadmap] Verified recovery objectives, replaceable infrastructure, mature fleet controls, assisted customer activation, and later self-service. ### Research [research] Additional deployment, identity, recovery, and agent-interface patterns are evaluated without being sold as available capability. --- ## Practical thinking for governed digital labor Canonical: https://aiemployeefoundry.com/insights Foundry insights help business owners distinguish a demo from an operating role, design human authority, choose deployment models, and measure useful work instead of AI activity. ### AI Employee vs. chatbot The six operating differences that turn conversation into an accountable role. ### Why authority comes first A practical framework for permissions, approvals, prohibitions, budgets, and emergency controls. ### Choosing a deployment When cloud, private, or hybrid infrastructure changes the operating model. --- ## AI Employee vs. chatbot: the operating difference Canonical: https://aiemployeefoundry.com/insights/ai-employee-vs-chatbot A chatbot primarily answers prompts. An AI Employee has a defined role, bounded authority, connected systems, recurring work, operating memory, human escalation, and measurable outcomes. ### Role A chatbot waits for a conversation. An AI Employee is defined by mission, owner, reporting line, scorecard, and recurring responsibilities. ### Authority A chatbot has features. An AI Employee has explicit permissions, approvals, prohibitions, spending limits, and escalation paths. ### Systems A chatbot may retrieve knowledge. An AI Employee works through scoped connections to the records and tools required for its job. ### Memory A chatbot remembers a thread. An AI Employee needs governed continuity: decisions, evidence, policies, work history, and approved institutional knowledge. ### Proof A chatbot is judged by response quality. An AI Employee is judged by work completed, exceptions handled, risks controlled, and business outcomes. ### Operation A chatbot is launched. An AI Employee is accepted, monitored, updated, supported, backed up, and eventually retired. --- ## Answers before deployment Canonical: https://aiemployeefoundry.com/faq AI Employee Foundry is Cloud Radix's managed system for defining, connecting, deploying, governing, and operating role-based AI Employees for real businesses. ### Is this a chatbot builder? No. Foundry begins with a business role and operating system, then uses the technologies that fit the approved work. ### How long does deployment take? Timing depends on workflow complexity, integrations, data sensitivity, authority design, and acceptance requirements. Discovery produces a responsible plan. ### Can it run on our hardware? Yes, when private or hybrid deployment is the right fit and the environment can meet the agreed network, access, support, and recovery requirements. ### Can we start small? Yes. A bounded recurring workflow with clear ownership and a measurable scorecard is often the strongest first role. ### Is pricing public? Not yet. Cloud Radix is finalizing repeatable packaging while continuing discovery-led implementations. ### Who supports it? Cloud Radix remains the provider, deployment partner, governance authority, and support relationship for Foundry implementations. --- ## Privacy at AI Employee Foundry Canonical: https://aiemployeefoundry.com/privacy This public website collects only the information a visitor chooses to submit for a strategy conversation and basic technical data needed to secure and operate the site. It does not expose or query customer AI Employee operational data. ### Information you submit Name, company, contact details, role or workflow interests, deployment preferences, and the context you choose to provide. ### How it is used To respond to the request, qualify the engagement, secure the service, and improve the public experience. Do not submit credentials, regulated records, case facts, or confidential client data. ### Operational separation Marketing leads and public content are separate from private Pistol Shrimp, Foundry runtime, customer, incident, backup, and secret data. ### Questions Contact skywalker@cloudradix.com about privacy requests. Formal contractual and customer data terms are established during an engagement. --- ## Clear expectations for this public website Canonical: https://aiemployeefoundry.com/terms This website provides general product and educational information. It is not a promise that every described roadmap feature is generally available, and it is not legal, financial, security, or professional advice. ### Maturity labels matter Available, managed pilot, roadmap, and research labels describe different levels of commercial readiness. ### No blanket guarantee Specific capability, service levels, responsibilities, data handling, and outcomes are defined in the applicable written customer agreement. ### Intellectual property AI Employee Foundry, Pistol Shrimp AI, designs, software, and original content remain the property of their respective owners. ### Responsible use Do not attempt to disrupt the site, bypass controls, submit unlawful material, or probe private systems through public endpoints. --- ## A site designed for people, not just crawlers Canonical: https://aiemployeefoundry.com/accessibility AI Employee Foundry targets a tested WCAG 2.2 Level AA conformance posture through semantic structure, keyboard operation, visible focus, reduced motion, contrast, responsive layouts, persistent form labels, and automated accessibility checks. ### Interaction Navigation and forms are designed for keyboard use, focus visibility, focus restoration, Escape behavior, and 44-pixel touch targets. ### Presentation Status is never communicated by color alone. Motion respects user preferences. Text and controls are designed for readable contrast and zoom. ### Testing Automated Axe and browser checks support—but do not replace—manual keyboard, zoom, contrast, and assistive-technology review. ### Feedback If you encounter a barrier, email skywalker@cloudradix.com with the page and task. We will investigate and improve it.