Skip to main content
Introducing TSK-1Introducing TSK-1·Taskade's intelligence layer.
taskade
Create
Connect
Clone
PricingHelpDashboard →Dashboard →
PricingLoginSign up for free →Sign up for free →
Dashboard →Dashboard →
Sign up for free →Sign up for free →
Loved by 1M+ users·Hosting 150,000+ apps·Deploying 500K+ AI agents·Running 1M+ automations·Backed by Y Combinator·Powered by TSK-1
Taskade
AppsTSK-1AI ModelsBuilderPricingFeatures
See all▾
Contact usIntegrationsMCP ServerAgent skillllms.txtPressAboutConnectProductivityVideosReviewsFAQ
Learn
Taskade GenesisProjectsAI AgentsAutomationConnectorsVideo Tutorials
See all▾
Import & ExportSearch ArticlesAccount & Billing
App Factory
FeaturedDashboardsOperations
See all▾
CommerceFinanceSalesCRMsInventoryPortalsInvoicesBookingToolsCreators
Downloads
Mobile appsDesktop appsAndroid
See all▾
iOSMacWindowsChromeFirefoxEdge
Compare
vs Cursorvs Boltvs Lovable
See all▾
vs V0vs Windsurfvs Replitvs Emergentvs Devinvs Claude Codevs ChatGPTvs Claudevs Perplexityvs GitHub Copilotvs Figma AIvs Notionvs ClickUpvs Asanavs Mondayvs Trellovs Jiravs Linearvs Todoistvs Evernotevs Obsidianvs Airtablevs Basecampvs Mirovs Slackvs Bubblevs Retoolvs Webflowvs Framervs Softrvs Glidevs FlutterFlowvs Base44vs Adalovs Durablevs Gammavs Squarespacevs WordPressvs UI Bakeryvs Zapiervs Makevs n8nvs Jaspervs Copy.aivs Writervs Rytrvs Manusvs Crewvs Lindyvs Relevance AIvs Wrikevs Smartsheetvs Monday Magicvs Codavs TickTickvs Any.dovs Thingsvs OmniFocusvs MeisterTaskvs Teamworkvs Workfrontvs Bitrix24vs Process Streetvs Toggl Planvs Motionvs Momentumvs Habiticavs Zenkitvs Google Docsvs Google Keepvs Google Tasksvs Microsoft Teamsvs Dropbox Papervs Quipvs Roam Researchvs Logseqvs Memvs WorkFlowyvs Dynalistvs XMindvs Whimsicalvs Zoomvs Remember The Milkvs Wunderlist
AI Agents
FeaturedProject ManagementProductivity
See all▾
MarketingTranslatorContentWorkflowResearchPersonalOperations IntelligenceSalesSocial MediaTo-Do ListCRMTask AutomationCoachingCreativityTask ManagementBrandingFinanceLearning and DevelopmentBusinessCommunity ManagementMeetingsAnalyticsDigital AdvertisingContent CurationKnowledge ManagementProduct DevelopmentPublic RelationsProgrammingHuman ResourcesE-CommerceEducationLegalEmailSEODeveloperVideo ProductionDesignFlowchartDataPromptNonprofitAssistantsTeamsCustomer ServiceTrainingTravel PlanningUML DiagramER DiagramMath TutorLanguage LearningCode ReviewerLogo DesignerUI WireframeFitness CoachLead EnrichmentFounder OSSales DevelopmentBookkeepingRecruitingWebsite MonitoringField ServiceLicensingAll Categories
Automations
FeaturedAI Workflow AutomationLogic Automations
See all▾
AI Agent AutomationTrigger AutomationsAgentic Process AutomationAction AutomationsAI Models in WorkflowsAgentic AutomationMulti-Agent AutomationSocial MediaBusiness-in-a-BoxOperations IntelligenceInvestor OperationsEducation & LearningHealthcare & ClinicsReal EstateStripeSalesHR & People OpsField Service & DispatchRenewals & LicensesE-commerceContentMarketingEmailCustomer SupportHubSpotProject ManagementAgentic WorkflowsAppointment SchedulingCalendarReportsSlackWebsiteFormTaskWeb ScrapingWeb SearchChatGPTText to ActionYouTubeLinkedInX (Twitter)GitHubDiscordMicrosoft TeamsWebflowIndustry News & RSS FeedsGoogle WorkspaceManufacturing & OperationsAI Agent TeamsNotion AutomationsProposalBookkeeping & ExpensesClient OnboardingGoogle SheetsGoogle DriveGoogle CalendarGoogle FormsShopifyAsanaAirtableTrelloTodoistMailchimpClickUpGoogle DocsGmailGoogle TasksJiraLinearMicrosoft OutlookTelegramCalendlyTypeformSalesforceMonday.comTwilioWhatsApp BusinessWordPressZoomHTTP and WebhooksMCP AutomationFacebook PagesRedditApolloPipedriveFigmaAttioBrevoCal.comConfluenceHighLeveliCal FeedJotformOutlook CalendarMicrosoft To DoPushoverResendSendGridSMTPSupabaseTallyAll Categories
Prompts
Client PortalsDocument IntakeField Service
See all▾
Operations IntelligenceCRMAdvertisingDesignInvoicing and PaymentsSocial MediaQuotes and EstimatesCalculatorsCodingBooking and SchedulingConsultingCustomer SupportComplianceTime TrackingRetail OperationsSaaS and SubscriptionsInsuranceGovernance and OversightNonprofitPatient IntakeBlog WritingBrandingPersonal FinanceHuman ResourcesPublic RelationsTeam CollaborationProduct ManagementSupportAgencyReal EstateMarketingResearchSalesCopywritingContentProject ManagementWebsite CreationStrategyE-commerceEngineeringSEOEducationEmail MarketingUX/UIProductivityInfluencer MarketingAnalyticsEntrepreneurshipLegalVibe CodingRecruitingAll Categories
Templates
FeaturedChatGPTTable
See all▾
PersonalProject ManagementSalesFlowchartTask ManagementEngineeringOperations IntelligenceInventory ManagementEducationDesignTo-Do ListMarketingMind MapGantt ChartOrganizationalPlanningMeetingsTeam ManagementStrategyGamingProductionProduct ManagementStartupRemote WorkY CombinatorRoadmapCustomer ServiceLegalEmailBudgetsContentConsultingE-CommerceStandard Operating Procedure (SOP)Human ResourcesProgrammingMaintenanceCoachingSocial MediaHow-TosResearchMusicTrip PlanningCRMClient OnboardingEmployee OnboardingSOPBug TrackerRecruitment TrackerFormSales PipelineContent CalendarMarketing PlanProduct RoadmapBusiness PlanSWOT Analysis30-60-90 Day PlanInterviewNotion AlternativeKPIStrategic PlanMeeting AgendaInvoiceRisk RegisterIT Asset ManagementKanban BoardChange ManagementCommunication PlanRFPScope of WorkStatement of WorkHelpdeskKnowledge BaseCreative BriefGoal SettingExecutive SummaryGap AnalysisBooking SystemEvent ManagementPortfolio TrackerCustomer Onboarding PortalsClient PortalAgency OperationsFinance TrackingAll Categories
Generators
AI SoftwareNo-Code AI AppAI App
See all▾
AI WebsiteAI DashboardAI FinanceAI Operations IntelligenceAI FormAI AgentAI Client Portal BuilderAI WorkspaceAI ProductivityAI To-Do ListAI WorkflowsAI EducationAI Mind MapsAI FlowchartAI Scrum Project ManagementAI Agile Project ManagementAI MarketingAI Project ManagementAI Social Media ManagementAI BloggingAI Agency WorkflowsAI ContentAI Software DevelopmentAI MeetingAI PersonasAI Outline GeneratorAI SalesAI ProgrammingAI DesignAI FreelancingAI ResumeAI Human ResourceAI SOPAI E-CommerceAI EmailAI Public RelationsAI InfluencersAI Content CreatorsAI Customer ServiceAI BusinessAI PromptsAI Tool BuilderAI SEOAI Gantt ChartAI CalendarsAI BoardAI TableAI ResearchAI LegalAI ProposalAI Video ProductionAI Health and WellnessAI WritingAI PublishingAI NonprofitAI DataAI Event PlanningAI Game DevelopmentAI Project Management AgentAI Productivity AgentAI Marketing AgentAI Personal AgentAI Business and Work AgentAI Education and Learning AgentAI Task Management AgentAI Customer Relations AgentAI Programming AgentAI SchemaAI Business PlanAI Pitch DeckAI InvoiceAI Lesson PlanAI Social Media CalendarAI API DocumentationAI Database SchemaAI Marketing PlanAI Sales Pipeline GeneratorAI Course BuilderInternal ToolsBooking SystemReal Estate CRMInventory ManagementAI CRM BuilderAI TimesheetAI DispatchAI NewsletterAI Clinic OperationsAI Directory BuilderAll Categories
Taskade AI
Video GuideAI App BuilderVibe Coding
See all▾
Agent BuilderDashboard BuilderCRM BuilderWebsite BuilderForm BuilderWorkflow AutomationWorkflow BuilderBusiness-in-a-BoxAI for MarketingAI for Developers
Converters
AI Featured ConvertersAI PDF ConvertersAI CSV Converters
See all▾
AI Markdown ConvertersAI Prompt to App ConvertersAI Data to Dashboard ConvertersAI Workflow to App ConvertersAI Idea to App ConvertersAI Flowcharts ConvertersAI Mind Map ConvertersAI Text ConvertersAI YouTube ConvertersAI Knowledge ConvertersAI Spreadsheet ConvertersAI Email ConvertersAI Web Page ConvertersAI Video ConvertersAI Coding ConvertersAI Task ConvertersAI Kanban Board ConvertersAI Notes ConvertersAI Education ConvertersAI Language TranslatorsAI Business → Backend App ConvertersAI File → App ConvertersAI SOP → Workflow App ConvertersAI Portal → App ConvertersAI Form → App ConvertersAI Schedule → Booking App ConvertersAI Metrics → Dashboard ConvertersAI Game → Playable App ConvertersAI Catalog → Directory App ConvertersAI Creative → Studio App ConvertersAI Agent → Agent App ConvertersAI Audio ConvertersAI DOCX ConvertersAI EPUB ConvertersAI Image ConvertersAI Resume & Career ConvertersAI Presentation ConvertersAI PDF to Spreadsheet ConvertersAI PDF to Database ConvertersAI PDF to Quiz ConvertersAI Image to Notes ConvertersAI Audio to Notes ConvertersAI Email to Tasks ConvertersAI CSV to Dashboard ConvertersAI YouTube to Flashcards ConvertersURL to NotesVideo → SummaryAI Receipts to Expense Tracker ConvertersAI Docs to Knowledge Base ConvertersAI Form to Client Portal ConvertersSpreadsheet to CRMAll Categories
Wiki
AI Agent ExamplesLiving DNAAutonomous Workspaces, Agents & Apps
See all▾
Taskade GenesisAI AgentsAutomationProjectsQuantum AI & Taskade Genesis QuantumFoundations: The Theory Under Modern AICognitive Science: Mind, Brain and AIPlatformAI InfrastructureIntegrationsProductivityMethodsProject ManagementAgileScrumAI ConceptsCommunityTerminologyFeatures
Docs
Getting StartedREST APIAction API
See all▾
MCP ServersGuides & SDKModels
Blog
Introducing Taskade App Factory (November 2026)How to Build an AI Support Agent for Your Website (2026)Residual Connections Explained: How ResNet's Skip Connections Made Deep Learning Deep (2026)
See all▾
AI Typography Tools 2026: 14 Font, Pairing and CSS ToolsWhy AI Agents Need an Ontology: The Write Path Nobody Talks About (2026)Browser Agents Explained: How AI Agents Use the Web in 2026Your Business. Your Software. (November 2026)Open Knowledge Format (OKF) Explained (2026): Every Rule in Google's Spec for Agent-Readable KnowledgeHistory of Microsoft Copilot: Clippy to Agents (2026)Is Taste the New Moat? What Still Defends Software When AI Builds Everything (2026)How to Build a Safety and Compliance Tracker Without Code (2026)History of Clay.com: From a Spreadsheet to a $7.1B GTM Platform (2017 to 2026)What Is x402? The HTTP 402 Payment Protocol for AI Agents, Explained (2026)How to Build an AI Lead Qualification Agent (No Code, 2026)LLM Wiki Explained (2026): Karpathy's Pattern, Google's OKF, and When a Folder Beats RAGWhat Is Spec-Driven Development? Tools and Evidence (2026)The Workspace Is the Runtime: What AI Agents Need After the Prompt (2026)Best AI Story Games 2026: 20 Tools to Play, Write, BuildHow to Build an Inventory Tracker Without Code (2026 Guide)Introducing Taskade TSK-1: The System Kernel Behind Every App (2026)Run Your Whole Business in One App with Taskade Genesis (June 2026)
AIAutomationProductivityProject ManagementRemote WorkStartupsKnowledge ManagementCollaborative WorkUpdates
Changelog
New Automation Triggers & Row Totals in Taskade EVE (Nov 16, 2026)Read PDFs in Taskade & File Previews in Chat (Nov 14, 2026)Approve From Activity & Steadier App Kit Copies (Nov 13, 2026)
See all▾
Locked List Alerts & Safer Published Apps (Nov 12, 2026)Automation Setup Checklist & PayPal Signature Checks (Nov 11, 2026)One-Click App Connections & Date Fields Over MCP (Nov 10, 2026)Public Agent Chat Privacy & Cleaner Kit Copies (Nov 9, 2026)
Integrations
Slack AutomationGmail AutomationGoogle Sheets Automation
See all▾
Google Calendar AutomationGoogle Drive AutomationHubSpot AutomationStripe AutomationDiscord AutomationGitHub AutomationNotionYouTube AutomationZapier
© 2026 TASKCADE INC.
PrivacyTermsSecurity
Made withTaskade AIforBuilders
BlogAIWhat Is an AI Employee?…

What Is an AI Employee? Roles, Costs, Risks and How to Hire Your First One (2026)

An AI employee is an AI agent with a role, memory, a schedule and a manager. See what one does, what it costs, what goes wrong and how to hire your first one.

Human computers at work in the NACA High-Speed Flight Station computer room in 1949, when computer was a job title. Photo by NACA and NASA, public domain, via Wikimedia Commons
November 19, 202633 min readJohn XieAI#ai-employees#ai-agents#small-business
On this page (20)
⚡ Quick Answer: What Is an AI Employee?🧮 Where Did the Idea of an AI Employee Come From?🆚 AI Employee vs AI Agent vs Chatbot vs Automation vs RPA Bot🧬 What Is an AI Employee Made Of?How do you give an AI employee an identity?🔁 How Does an AI Employee Work Day to Day?📊 How Many Small Businesses Use AI in 2026?Why do the surveys disagree so much?🏪 Who Sells AI Employees in 2026?💵 What Does an AI Employee Cost to Run?⚠️ What Goes Wrong with AI Employees?🎯 Which Roles Should You Hire First?🤝 What Is a Managed AI Employee, and Is It a Business?🧭 How Do You Hire Your First AI Employee?📝 A One-Page Job Description for an AI Employee🧬 Where Does Taskade Genesis Fit?✅ Is Your Business Ready? A 10-Question Checklist🔑 Key Takeaways💬 Frequently Asked Questions🔗 Further Reading

In the summer of 1949, a photographer walked into a room at the NACA High-Speed Flight Station in California and took a picture of the computers. There was no electronic machine in the frame. The computers were people at desks with slide rules, electric calculators, rolls of film and strips of oscillograph paper. NASA's own caption for the photo opens with the point: "In the terminology of that period, computers were employees" who reduced raw flight data to engineering units by hand.

Then the word moved. Within a generation, "computer" stopped meaning a person and started meaning the tool that did the arithmetic. The people did not vanish. They became programmers, engineers and analysts, and they managed the machines that had taken over their old job title.

In 2026 the same thing is happening to another word. Vendors now sell an "AI employee" for the price of a streaming bundle. This guide explains what that phrase really means, what the data says about who uses one, what it costs, what goes wrong, and how to hire your first one without losing control of your business.

TL;DR: An AI employee is an AI agent packaged as a job: a role, knowledge, tools, memory, a schedule and a human manager. Goldman Sachs found that 76% of small businesses use AI and 14% fully integrate it. The gap is setup, not intelligence. Start with one low-risk role and an approval gate. Build one in Taskade Genesis →

Photo: human computers in the NACA High-Speed Flight Station computer room, 1949. NACA and NASA, public domain, via Wikimedia Commons.

⚡ Quick Answer: What Is an AI Employee?

An AI employee is an AI agent that has been given a job. It has a role with a written job description, the knowledge and tools that role needs, a memory that carries over between runs, a schedule that starts the work without a prompt, and a human manager who approves the result. The model is rented. The job design is yours.

Part Plain meaning Without it
Role One job with a clear outcome A general assistant that does a little of everything
Knowledge Your documents, customers and rules Generic answers
Tools The apps it can read and act in Advice with no action
Memory What it keeps between runs It starts from zero each time
Schedule A clock or trigger that starts the work You have to ask every time
Manager A named person who approves Nobody owns the result
AI modelpredicts the next words AI agenta model that uses tools in a loop AI employeean agent with a role, memory, schedule and manager Managed AI employeean operator runs it for a monthly fee
AI modelpredicts the next words AI agenta model that uses tools in a loop AI employeean agent with a role, memory, schedule and manager Managed AI employeean operator runs it for a monthly fee

The picture to keep in mind is a staffing ladder. Each step up adds packaging, not intelligence. That matters when you compare products, because two "AI employees" can run on the same model and behave very differently.

🧮 Where Did the Idea of an AI Employee Come From?

The idea is older than the technology. For most of the twentieth century a "computer" was a person, and the word was a job title. NASA's history of the practice is blunt about it: "the term 'computer' referred to people, not machines." The job title moved to the machine once the machine could do the job.

Langley formed its first computer pool in 1935 with five women. A 1942 report counted 75. By 1946 the department had trained about 400. They read raw data from film and paper, then reduced it to engineering units by hand. When electronic machines arrived, many of those same people learned to program them.

Year What "computer" or "employee" meant Source
1935 Langley forms its first computer pool: five women who calculate by hand NASA history
1942 75 women work as computers at Langley Same
1949 The photo above: computers at desks with slide rules and calculators NACA photo caption
1950s to 1960s Electronic machines take the arithmetic. The word follows the work History of computing
2025 to 2026 Vendors sell "AI employees" and "digital workers" by the month Vendor table below
1935: computer is a job titlea person who calculates 1950s: machines take the arithmeticthe word moves to the tool The people move upprogrammers, engineers, analysts 2026: employee is becoming a rolethat software can hold The people move up againmanagers of AI agents
1935: computer is a job titlea person who calculates 1950s: machines take the arithmeticthe word moves to the tool The people move upprogrammers, engineers, analysts 2026: employee is becoming a rolethat software can hold The people move up againmanagers of AI agents

The lesson from 1949 is not that machines replace people. It is that a job title names a bundle of tasks, and the bundle can be reassigned. What stayed with the people was judgment: deciding what to calculate and whether the answer made sense. That is the part to keep when you hire an AI employee. For the longer arc, see the history of AI agents.

🆚 AI Employee vs AI Agent vs Chatbot vs Automation vs RPA Bot

An AI employee differs from the other four in one way: it owns a recurring outcome and reports to a person. A chatbot answers when asked. An automation follows fixed steps. An RPA bot repeats clicks on a screen. An agent pursues a goal with tools. An AI employee is an agent that has a job, a memory and a manager.

Term What it does What starts it Handles the unexpected?
Chatbot Answers questions in a conversation A person's message Only in words
Automation Runs fixed steps in order A trigger or schedule No, it stops or errors
RPA bot Repeats clicks and keystrokes in a user interface A schedule No, it breaks when the screen changes
AI agent Uses tools in a loop to reach a goal A prompt or trigger Yes, within its tools
AI employee Owns a recurring outcome in a defined role A schedule, with a manager's review Yes, and it asks before risky actions
Yes No Yes No No Yes Does it only answer questions? Chatbot Does it follow fixed steps every time? Automation or RPA bot Does it have a role, a memory and a manager? AI agent AI employee
Yes No Yes No No Yes Does it only answer questions? Chatbot Does it follow fixed steps every time? Automation or RPA bot Does it have a role, a memory and a manager? AI agent AI employee

These categories overlap in practice. A good AI employee usually contains an automation (the schedule and the fixed steps) and an agent (the judgment in the middle). If you want the full map of terms, we keep one in the AI agents taxonomy, and a plain primer in what AI agents are. The difference between an assistant that suggests and a worker that acts is covered in agents versus copilots.

One caution about the name. "Employee" is a metaphor. The software has no contract, no rights and no accountability. The accountability stays with the person who manages it, which is why the manager appears in the definition.

🧬 What Is an AI Employee Made Of?

An AI employee has seven parts: a role, knowledge, tools, memory, a schedule, an approval gate and an audit trail. The first four make it capable. The last three make it safe to leave running. Most disappointing deployments have the first four and skip the last three.

Part Question it answers Example for a follow-up role
Role What is this job, and what is it not? Draft follow-ups for leads with no reply in 5 days
Knowledge What does it need to know about us? Pricing sheet, tone guide, past winning emails
Tools Where can it read and act? CRM table, email, calendar
Memory What does it keep? Each lead's history and your past edits
Schedule When does it work? Every weekday at 8:00
Approval gate What needs a human yes? Every email before it is sent
Audit trail What did it do, and who approved? A log row for each draft and decision
Roleone job, one outcome Knowledgeyour documents and rules Toolsthe apps it can use Memorywhat it keeps between runs Schedulestarts the work without a prompt Approval gatea named person says yes or no Audit trailevery action and decision is recorded
Roleone job, one outcome Knowledgeyour documents and rules Toolsthe apps it can use Memorywhat it keeps between runs Schedulestarts the work without a prompt Approval gatea named person says yes or no Audit trailevery action and decision is recorded

How do you give an AI employee an identity?

Write it down in plain files. The clearest public pattern comes from the open-source agent project OpenClaw, whose workspace convention splits identity into short Markdown files that the agent reads at the start of every session. You do not need that project to borrow the idea.

File What it holds Human equivalent
USER.md Who you are and how to address you Knowing your boss
SOUL.md Persona, tone and boundaries Company culture
IDENTITY.md The agent's name, vibe and emoji The name on the badge
AGENTS.md Operating rules and how to use memory The employee handbook
MEMORY.md Long-term notes Experience

The format matters less than the habit. An AI employee with no written identity behaves like a temp on the first morning, every morning. The same idea, applied to a whole team's knowledge, is what we call context engineering.

🔁 How Does an AI Employee Work Day to Day?

An AI employee works in a loop: it reads the data, drafts the work, waits for a person, acts on the answer and records what happened. The waiting step is the important one. Engineers call this design human-in-the-loop, and it is the reason a business owner can trust a system that is sometimes wrong.

Schedule fires, agent reads the records Here is the draft and why Approve or reject, with a reason Send, post or update, only if approved Record the action and the decision Next draft starts from what was learned Your data AI employee Human manager Outside world Memory
Schedule fires, agent reads the records Here is the draft and why Approve or reject, with a reason Send, post or update, only if approved Record the action and the decision Next draft starts from what was learned Your data AI employee Human manager Outside world Memory

The LangChain agent framework documents the pattern with four decisions for a paused action: approve, edit, reject or respond. The table shows the three you will use most. Products differ in which decisions they offer, so check before you buy.

Decision What happens When to use it
Approve The action runs as drafted The draft is right
Edit You change the draft, then it runs The draft is close
Reject The action stops, and the reason is stored The draft is wrong or the timing is off

Notice what the loop does over time. Every approval and every rejection is a labelled example of your judgment. After a few months the system holds a record that no competitor has. We made the longer case for that record in is taste the new moat, and there is a step-by-step build in how to build an approval workflow.

📊 How Many Small Businesses Use AI in 2026?

Between 20% and 76% of businesses use AI, depending on who asks, whom they ask and how. The US Chamber of Commerce reports 66%. Goldman Sachs reports 76%. The Census Bureau, which asks about the last two weeks only, reports about 20%. The spread is the story: most firms have tried AI, and few run it every day.

Survey Finding Date Source
US Chamber of Commerce 66% of small businesses use AI, up from 58% in 2025 and 23% in 2023 Oct 2, 2026 US Chamber
Goldman Sachs 10,000 Small Businesses 76% use AI, and 14% have fully integrated it Jan to Feb 2026 Goldman Sachs
Salesforce SMB Trends 75% are experimenting with or using AI 6th edition Salesforce
Census Bureau BTOS 19.8% of firms used AI in the two weeks to May 3, 2026 May 2026 Census Bureau
Census BTOS US Chamber Salesforce Goldman Sachs 0 20 40 60 80 100 Businesses using AI (%) Share of businesses that say they use AI, by survey (%)
Census BTOS US Chamber Salesforce Goldman Sachs 0 20 40 60 80 100 Businesses using AI (%) Share of businesses that say they use AI, by survey (%)

Why do the surveys disagree so much?

They measure different things. A survey that asks "do you use AI?" counts an owner who opened a chat assistant once. A survey that asks whether the firm used AI in the last two weeks counts only regular use. The Census Bureau also broadened its question in November 2025, so its older figures do not compare with the new ones.

Reason Effect on the number
Question wording: "use" versus "used in the last two weeks" The loose wording gives a much higher result
Who is asked: program alumni versus a national sample Engaged owners report higher use
Firm size: all employer firms versus small businesses with staff Very small firms pull the national figure down
Depth: tried it versus built it into operations Goldman Sachs: 76% use, 14% fully integrate
2023 2025 2026 0 20 40 60 80 100 Businesses using AI (%) US Chamber of Commerce: small businesses that use AI (%)
2023 2025 2026 0 20 40 60 80 100 Businesses using AI (%) US Chamber of Commerce: small businesses that use AI (%)

The number that matters for this guide is the Goldman Sachs pair: 76% use AI and 14% have fully integrated it. That 62-point gap is the market for AI employees. The owners in the gap do not need a smarter model. They need the packaging from the anatomy section: a role, a place for the data, a schedule and a gate. A separate survey found little appetite for layoffs. In the US Chamber data, 47% of small businesses said AI creates jobs and 6% said it enables cuts.

🏪 Who Sells AI Employees in 2026?

Three kinds of company sell AI employees in 2026: role catalogs that offer prebuilt helpers, builder platforms where you assemble your own, and managed services where an operator runs the system for you. Prices run from about $24 a month for a self-serve catalog to thousands a month for enterprise sales agents.

The table below lists eight vendors and the latest price we could find for each. Prices are as of November 2026 and change often, so check the live page before you decide.

Vendor What it sells Price found Price source As of
Sintra A catalog of role-based helpers for small businesses $39 a month per helper, or a bundle of 12 at $97 list Toolradar Nov 2026
Lindy A builder for assistants that work across email, calendar and apps Free trial, Team from $29.99 a user a month with 3,000 credits, Enterprise custom Lindy Nov 2026
Marblism Seven prebuilt AI employee roles $24 a month yearly, $44 monthly Marblism Nov 2026
Viktor An AI coworker for teams, billed by credits Free starting credits, Team at $100 a month Toolworthy Nov 2026
Relevance AI A platform for building an AI workforce Enterprise on request on its own page. A directory lists Pro $29 and Team $349 ReachKit Nov 2026
Devin An AI software engineer Free, Pro $20, Max $200, Teams from $80 Replicas Nov 2026
11x AI sales development representatives Outbound from $36,000 a year, Inbound from $2,417 a month, billed annually 11x Nov 2026
Artisan An AI sales representative By sales quote The CRO Report Nov 2026

Two strengths are worth naming, because they explain why the category grew so fast.

  • Role catalogs made the idea legible. Sintra and Marblism let an owner pick a role from a menu and start the same day. That is real product craft: a role is easier to understand than a workflow builder.
  • Builders made the idea flexible. Lindy and Relevance AI let a team shape an assistant around the specific apps and data it already uses. That is the difference between a demo and a system.
You want recurring work done Role catalogpick a prebuilt helper Builder platformassemble your own Managed servicean operator runs it for you Fastest startleast control over the job Best fit to your toolsyou do the setup Least efforthighest monthly price
You want recurring work done Role catalogpick a prebuilt helper Builder platformassemble your own Managed servicean operator runs it for you Fastest startleast control over the job Best fit to your toolsyou do the setup Least efforthighest monthly price
Delivery model You get You give up Best for
Role catalog A working helper today Fit to your exact process A first experiment
Builder platform A system shaped to your tools and data Setup time A team with a clear process
Managed service An outcome each month Price and some visibility An owner with no time to build

One buying note. Independent data on how well these products retain customers is thin. We could not find a third-party retention study for any AI employee vendor. Treat vendor case studies as marketing, run a paid pilot on one role, and measure it yourself. For a deeper look at one open-source option, see our Suna review, and for the broader field, AI agent platforms compared.

💵 What Does an AI Employee Cost to Run?

An AI employee you build yourself costs four things each month: runtime, model tokens, the tools it connects to, and your own review time. Runtime is the smallest. Anthropic lists Claude Managed Agents at "$0.08 USD per session-hour", billed only while a session is running, with tokens charged separately at standard rates.

Cost part What drives it How to keep it down
Runtime Hours the agent is actually running Short, scheduled sessions instead of always-on
Tokens How much it reads and writes each run One place for data, so it reads less to find more
Tools Subscriptions for the apps it touches Start with the apps you already pay for
Review time Minutes you spend on each approval Clear drafts with the reason attached
Runtimesession-hours while running Tokenswhat the model reads and writes Tools and integrationsthe apps it connects to Human reviewyour minutes per approval True monthly costcompare it with the hours returned
Runtimesession-hours while running Tokenswhat the model reads and writes Tools and integrationsthe apps it connects to Human reviewyour minutes per approval True monthly costcompare it with the hours returned

Here is one worked example. The assumptions are ours, so change them to match your case.

WORKED EXAMPLE: ONE FOLLOW-UP ROLE, ONE MONTH

Assumptions
- 2 scheduled sessions each working day
- 30 minutes of running time per session
- 22 working days in the month

Runtime
- 2 x 0.5 hours x 22 days       = 22 session-hours
- 22 session-hours x $0.08      = $1.76 for the month

Not included above
- Tokens: billed separately at the model's standard rates
- Tools: your existing email and CRM subscriptions
- Review: about 2 minutes per draft, for every draft you approve

What to compare it with
- The hours of follow-up work you no longer do by hand

Two honest caveats. First, the runtime figure is the smallest line, and tokens usually cost more than runtime. Price them from the rate card of the model you use. Second, your review time is real cost. A role that produces 40 drafts a day and needs two minutes each has created a new 80-minute job for you. Design the role so that the drafts are few and good.

Packaged products fold all of this into one subscription and add their own margin for the convenience. That trade is often worth it for a first role. It is the same shift from seats to usage that we traced in the great SaaS unbundling.

⚠️ What Goes Wrong with AI Employees?

AI employees fail for ordinary reasons: scattered data, vague roles, no approval gate and no measure of success. The model is rarely the weak point. An agent that cannot find your customer's history will guess, and a guess sent under your name costs more than the hour it saved.

Failure mode Early sign Fix
The agent guesses Drafts that sound right and contain wrong facts Put the data in one place before you add the agent
Unreviewed actions A message went out that nobody read Add an approval gate in front of every outbound action
Vague role The agent does a little of everything, badly One role, one outcome, one page
Credit burn The monthly allowance is gone in the first week Shorter runs, smaller scope, a spending cap
Review overload You approve without reading Fewer, better drafts, and a reason on each one
No measure Nobody can say if it helps One number, with a baseline taken before the start
Guarantee without limits A promise of results you cannot control Guarantee the work delivered, not the client's revenue
Data is scatteredacross five apps The agent cannot find the facts It writes a confident draft anyway Approval gate in place No gate A person catches itthe correction is recorded The message goes outtrust is lost
Data is scatteredacross five apps The agent cannot find the facts It writes a confident draft anyway Approval gate in place No gate A person catches itthe correction is recorded The message goes outtrust is lost

Three of these deserve a closer look.

Missing context. The most common failure starts before the agent exists. Every business already has its data in a CRM, a spreadsheet, a booking app, an inbox and a folder nobody opens. An agent placed on top of that makes decisions with no idea of the context. Consolidate first. This is the same problem that agent memory research has circled for years.

Vendor claims. Check them. In March 2025, TechCrunch reported that the AI sales startup 11x had listed customer logos for companies that said they had only run short trials. The lesson is not about one vendor. In a young category, ask for a reference you can call and a pilot you can measure.

Outcome guarantees. Some operators sell AI employees with a 90-day results guarantee. A guarantee can win a client. It is also a liability you must price. Guarantee what you control, such as drafts delivered and response time, and leave the client's revenue out of the contract.

🎯 Which Roles Should You Hire First?

Hire for high volume, low risk and an easy measure. Reporting, research and inbox triage come first because a mistake costs little and a person sees the output before anyone else does. Roles that send messages come second. Roles that move money or make promises to clients come last.

Role Input Output Approval needed? How to measure
Weekly reporting Sales and project records A one-page summary No, it is internal Hours saved each week
Research A question and your sources A brief with links No, it is internal Briefs used without rework
Inbox triage Incoming email Labels, priorities, draft replies Yes, for replies Time to first reply
Lead follow-up Leads with no reply A personal follow-up draft Yes Reply rate
Support drafts Tickets and your help content A reply draft with a source Yes Share approved without edits
Content scheduling A calendar and approved posts Posts queued at the right time Yes, for new copy Posts shipped on time
Scheduling Calendars and booking rules Proposed times, confirmations Yes, for client meetings Back-and-forth messages saved
Invoice reminders Unpaid invoices A polite reminder draft Yes Days sales outstanding
Start hereinternal, low risk: reports, research Then drafts a person sendsfollow-ups, support replies, content Then actions with commitmentsclient scheduling, invoice reminders Last, and only with a record of accuracyanything that moves money
Start hereinternal, low risk: reports, research Then drafts a person sendsfollow-ups, support replies, content Then actions with commitmentsclient scheduling, invoice reminders Last, and only with a record of accuracyanything that moves money

A simple test picks the first role. Write down what you did last week that you will do again next week in almost the same way. The longest item on that list is your first hire. If you run alone, our guides to AI agents for solopreneurs and the solo operator business stack show the order that works for a one-person shop.

🤝 What Is a Managed AI Employee, and Is It a Business?

A managed AI employee is a service: an operator builds the system, runs it and charges a monthly fee for the outcome. It is the agency version of the product. For the client it removes setup. For the operator it replaces one-time build fees with recurring revenue, as long as the work keeps happening.

One-time build Managed AI employee
What the client buys A workflow An outcome that keeps happening
Payment Once Monthly
Who fixes it when it breaks Unclear The operator, by contract
What grows over time Nothing The record of the client's decisions
Reason to stay Low The system knows how they work

The market is large and easy to overstate, so here are the sourced numbers.

Segment Size Source
US franchise brands About 9,000 active brands FRANdata 2026 outlook
US franchise establishments About 845,000 forecast for 2026 International Franchise Association
Employer firms with $1M or more in receipts About 1.4 million in the 2014 Annual Survey of Entrepreneurs Census Bureau
Businesses with no employees and $1M or more in receipts 117,060 in 2023 Forbes, from Census data

You will see larger numbers quoted in videos, such as "30,000 franchise systems". The FRANdata count is about 9,000 brands. The opportunity is real without the inflation: hundreds of thousands of locations with steady cash flow and the same recurring problems.

What makes the service durable is not the technology, which any operator can buy. It is the memory. Every approval and rejection teaches the system the client's voice, priorities and preferences. Agency builder Luuk Alleman put it in a sentence: "The switching cost is the memory." A rival who wants the account has to earn six months of the client's decisions again.

Month 1data in one place, one role live Month 3hundreds of approvals and rejections recorded Month 6drafts match the client's voice Month 12leaving means teaching a new system from zero
Month 1data in one place, one role live Month 3hundreds of approvals and rejections recorded Month 6drafts match the client's voice Month 12leaving means teaching a new system from zero

A fair warning for operators: a monthly fee means monthly accountability. If the outcome stops, so does the revenue. That is the point of the model, and it is why a human approval gate protects the operator as much as the client. For the wider picture of what a single operator can run, see one-person companies.

🧭 How Do You Hire Your First AI Employee?

Hire your first AI employee in seven steps: pick one role, put the data in one place, write the job description, connect the tools, add the approval gate, set the schedule, and review every week. The order matters. Most failed attempts start at step four.

  1. Pick one role. Choose the recurring task that takes the most of your week.
    The longest item on last week's repeat list is the right one.
  2. Put the data in one place. Move the records the role needs into one workspace.
    An agent that reads five systems will miss the fact that matters.
  3. Write the job description. One page: role, inputs, tools, never-do list, approval rules, measure.
    The template is in the next section.
  4. Connect the tools. Give it the fewest apps the role needs, with the narrowest access.
    Read access first. Write access after the pilot.
  5. Add the approval gate. Nothing leaves the building without a named person's yes.
    Attach the reason to each draft so the review takes seconds.
  6. Set the schedule. Start with one run a day at a fixed time.
    A schedule is what turns an assistant you ask into an employee who shows up.
  7. Review every week. Read the approval rate, the measure and three rejected drafts.
    Change the job description, not the model, when the drafts are off.

Trust should grow in stages, and each stage should be earned by a number.

drafts are accurate for two weeks drafts are wrong too often high approval rate for a month the measure does not move Pilot: drafts only, you read everything Supervised: acts after your approval Trusted for low-risk actions Paused: fix the data or the role
drafts are accurate for two weeks drafts are wrong too often high approval rate for a month the measure does not move Pilot: drafts only, you read everything Supervised: acts after your approval Trusted for low-risk actions Paused: fix the data or the role
Stage What the AI employee may do What moves it forward
Pilot Draft only. Nothing is sent Two weeks of accurate drafts
Supervised Act after each approval A month with a high approval rate and a better measure
Trusted for low-risk actions Act alone on internal, reversible work Stays here. Risky actions keep the gate

Keep the gate on anything that sends, posts or pays, however good the record looks. The cost of one bad message is larger than the minutes the gate takes.

📝 A One-Page Job Description for an AI Employee

A written job description is the cheapest control you have. It tells the agent what the job is, and it tells you what to check. Copy the template, fill it in for one role, and give the same page to the agent as its instructions and to the person who will manage it.

JOB DESCRIPTION: AI EMPLOYEE (one page)

ROLE
- Title: Lead follow-up assistant
- Outcome: every lead with no reply in 5 days gets a personal follow-up draft

INPUTS
- Leads table (name, company, last contact, notes)
- Pricing sheet and tone guide
- The 10 best follow-up emails we have sent

TOOLS
- Read: leads table, email history, calendar
- Write: draft emails only

NEVER DO
- Send anything without approval
- Quote a price that is not in the pricing sheet
- Contact a lead marked "do not contact"

APPROVAL RULES
- Every email: approval by the account owner
- No answer in 3 days: do not send, flag it in the weekly review

SUCCESS METRIC
- Reply rate on follow-ups, against last quarter's baseline
- Share of drafts approved without edits

REVIEW CADENCE
- Weekly: approval rate, the metric, 3 rejected drafts and why
- Monthly: update this page

The "never do" list earns its place faster than any other section. People and models both learn a boundary best from a clear statement of what is off limits. If you want a place to keep these pages and the knowledge they point to, our guide to AI second brain tools compares the options.

🧬 Where Does Taskade Genesis Fit?

Taskade Genesis fits the builder model: you assemble an AI employee from parts that live in one workspace. The workspace holds the data, an AI agent holds the role, an automation supplies the schedule and the approval step, and a published app gives clients a place to sign in. We call the combination Workspace DNA: Memory, Intelligence and Execution.

Part of an AI employee In Taskade Genesis Fact to know
One place for data Projects and databases in a workspace Taskade EVE keeps project memory in the workspace
Role and knowledge AI agents with custom instructions and knowledge Paste your one-page job description as the instructions
Tools and schedule Automations with 100+ integrations and a Schedule trigger Automations run on every plan. Free has 10 runs a billing period. Premium integrations need a paid plan
Approval gate The Wait for approval step It offers Approve and Reject. The run pauses until a teammate answers or the wait times out
Client access A published app with App User sign-in Sign-in confirms identity. You match records to the login and test with a second account
Workspaceprojects and databases hold the data AI agentinstructions and knowledge define the role Automationa schedule starts the work Wait for approvala teammate approves or rejects Published appclients sign in and see their work
Workspaceprojects and databases hold the data AI agentinstructions and knowledge define the role Automationa schedule starts the work Wait for approvala teammate approves or rejects Published appclients sign in and see their work

Three honest notes about our own product.

  • The approval step offers Approve and Reject. It has no edit decision inside the step, so keep the draft in a project where the approver can change it before approving.
  • Taskade routes frontier models from top AI labs and meters AI usage with credits, as most of the industry does. Our bet is the workspace around the model, described in how Workspace DNA works.
  • No product can write your job description for you. The role, the never-do list and the measure are your decisions.

To start, describe the role in one prompt and let Taskade Genesis build the table, the agent and the automation together. Pro is $10 a month billed annually. If you prefer to begin from a working example, clone one from AI apps or the Taskade App Factory.

✅ Is Your Business Ready? A 10-Question Checklist

A business is ready for an AI employee when it can answer yes to most of these ten questions. Score one point for each yes. The checklist takes five minutes and saves the month you would lose by starting at the wrong step.

# Question What it tests
1 Can you name one task you repeat every week in almost the same way? A real role
2 Is the data for that task in one place? Context
3 Could you explain the task to a new hire in one page? A job description
4 Do you know what the agent must never do? Boundaries
5 Is there a named person who will approve the output? Accountability
6 Can that person review a draft in under two minutes? Review load
7 Do you have one number that shows if it works? A measure
8 Did you record that number before you started? A baseline
9 Do you have a monthly spending limit in mind? Cost control
10 Will you read three rejected drafts every week? A feedback loop
Score Reading
8 to 10 Ready. Start the pilot this week
5 to 7 Close. Fix the data and the job description first
0 to 4 Not yet. Pick the role and consolidate the records before you buy anything

🔑 Key Takeaways

  1. An AI employee is packaging, not a new kind of intelligence. It is an agent with a role, knowledge, tools, memory, a schedule and a manager.
  2. The word is moving, as "computer" did. The job title goes to the software, and the people move up to manage it.
  3. Use is wide and shallow. 66% to 76% of small businesses say they use AI. In the Goldman Sachs survey, 14% have fully integrated it.
  4. Three ways to buy: a role catalog, a builder platform, or a managed service.
  5. Runtime is cheap, and review time is not. Design roles that produce few, good drafts.
  6. Failures are ordinary: scattered data, vague roles, no gate, no measure.
  7. Hire in order of risk. Internal work first, drafts second, money last.
  8. The approval record is the asset. It is what makes the system yours and hard to replace.

In 1949 the computers were employees. The people in that photo did not lose to the machine. They learned to direct it. The same offer is on the table now, with one condition: you have to write the job description.

Memory holds what your business knows. Intelligence applies it. Execution does the work. ▲ ■ ●

💬 Frequently Asked Questions

What is an AI employee?

An AI employee is an AI agent packaged as a job. It has a role with a job description, the knowledge and tools the role needs, a memory of past work, a schedule that starts the work without a prompt, and a human manager who approves what it does. The model supplies the intelligence. The packaging makes it behave like a member of staff.

What is the difference between an AI employee and an AI agent?

An AI agent is a model that uses tools in a loop to reach a goal. An AI employee is an agent with four additions: a defined role, a memory that persists between runs, a schedule or trigger so it works without being asked, and a manager who reviews its output. Every AI employee is an agent. Most agents are not set up as employees.

What can an AI employee do for a small business?

An AI employee handles recurring work that follows a pattern: lead follow-up, inbox triage, weekly reports, support reply drafts, scheduling, invoice reminders, content scheduling and research. It drafts the work and a person approves it. The best first roles have high volume, low risk and a clear measure.

How much does an AI employee cost?

Packaged AI employee products ranged from about $24 to $200 a month for self-serve plans in November 2026, and enterprise sales products started in the thousands. If you build your own, the cost has four parts: runtime, model tokens, the tools it connects to, and your review time. Anthropic lists Claude Managed Agents runtime at $0.08 per session-hour, with tokens billed separately.

Can I approve what an AI employee does before it acts?

Yes, and you should. The standard design is human-in-the-loop: the agent prepares an action, the run pauses, and a named person approves or rejects it. Use the gate in front of anything that sends a message, posts in public or moves money.

Do AI employees remember context?

They remember what the system stores for them. A useful AI employee reads from one place that holds your customers, projects and past decisions, and it writes its results back to the same place. The record of your approvals and rejections is the most valuable memory, because it teaches the system your preferences.

Which roles should I give an AI employee first?

Start with work that is frequent, low in risk and easy to measure: weekly reporting, research briefs and inbox triage. Move next to drafts that a person sends, such as lead follow-ups and support replies. Leave actions that involve money or commitments to clients until the drafts have been accurate for at least a month.

How many small businesses use AI in 2026?

It depends on the survey. The US Chamber of Commerce reported in October 2026 that 66% of small businesses use AI. Goldman Sachs found 76% using AI and 14% fully integrating it. The Census Bureau, which asks about the last two weeks, measured about 20% in May 2026. Trying AI is common. Running it in daily operations is not.

Do AI employees replace human employees?

In most small businesses they change the job more than they remove it. The person who did the recurring work becomes the manager who reviews it. In the US Chamber of Commerce survey of October 2026, 47% of small businesses said AI creates jobs and 6% said it enables cuts.

How do I measure the ROI of an AI employee?

Pick one number before you start and record the baseline. Good measures are hours saved each week, time to first reply, the share of drafts approved without edits, and overdue invoices. Compare the monthly cost, including your review time, with the value of the hours returned.

Can one person run a company with AI employees?

One person can run more than before, with limits. The Census Bureau counted 117,060 businesses with no employees and at least $1 million in receipts in 2023. AI employees extend what a single operator can supervise, but each role still needs a source of data, a written job description, an approval rule and a weekly review.

🔗 Further Reading

  • The Agent Harness Explained: the software around the model that makes an agent reliable
  • Best AI Agent Builders: tools for assembling your own agents
  • The SaaSpocalypse Explained: why software pricing moved from seats toward work done

Build this in Taskade

Open this prompt in Taskade Genesis, edit it, and run it when it reads right:

Build an AI employee workspace: a table of leads with status and last contact date, an AI agent that drafts a follow-up for each stalled lead, and an automation that waits for my approval before it sends the email.

Start with this prompt

View this page as Markdown

0%

On this page

⚡ Quick Answer: What Is an AI Employee?🧮 Where Did the Idea of an AI Employee Come From?🆚 AI Employee vs AI Agent vs Chatbot vs Automation vs RPA Bot🧬 What Is an AI Employee Made Of?How do you give an AI employee an identity?🔁 How Does an AI Employee Work Day to Day?📊 How Many Small Businesses Use AI in 2026?Why do the surveys disagree so much?🏪 Who Sells AI Employees in 2026?💵 What Does an AI Employee Cost to Run?⚠️ What Goes Wrong with AI Employees?🎯 Which Roles Should You Hire First?🤝 What Is a Managed AI Employee, and Is It a Business?🧭 How Do You Hire Your First AI Employee?📝 A One-Page Job Description for an AI Employee🧬 Where Does Taskade Genesis Fit?✅ Is Your Business Ready? A 10-Question Checklist🔑 Key Takeaways💬 Frequently Asked Questions🔗 Further Reading

Related Articles

How to build an app factory: a cloned Taskade App Kit with its projects, AI agent and automations in the side panel, and Taskade EVE ready to make it your own
November 5, 2026AI

How to Build an App Factory Without Code: The 2026 Guide

An app factory gives you a few apps that work your way, with no build from scratch. Clone an App Kit, give it a brain, m...

AI software factory for teams who do not code: a Taskade Genesis task dashboard that tracks requests from To Do to Done
October 25, 2026AI

What Is an AI Software Factory? And How to Run One Without Code (2026)

What an AI software factory is, where it came from, how the 2026 loop works, how to measure it, and how a team that does...

How to build an AI support agent for your website: a row of Chesapeake and Potomac Telephone Company operators at a switchboard in 1927, an early human handoff desk. Photo: U.S. National Archives (NARA 522875) via Wikimedia Commons, public domain
November 19, 2026AI

How to Build an AI Support Agent for Your Website (2026)

Put an AI support agent on your website that answers from your docs, hands hard questions to a person, and logs each han...

How to build an AI lead qualification agent without code: the Lead Score Desk kit in Taskade Genesis, with a hot-lead count, fit and readiness scores and a fit check form. Screenshot of a public Taskade App Factory kit, captured 2026-11-07.
November 14, 2026AI

How to Build an AI Lead Qualification Agent (No Code, 2026)

Build an AI agent that scores every new lead from 0 to 100, routes hot leads to an owner, and follows up with quiet ones...

Spec-driven development: the Taskade sheep mascot thinks over an open notebook with a pencil and a ruler before it builds
November 13, 2026AI

What Is Spec-Driven Development? Tools and Evidence (2026)

Spec-driven development means you write what the software must do before an AI agent builds it. The loop, the tools, the...

AI agent sandboxes explained: a wooden sandbox with toys inside a clear boundary, the physical version of where an AI agent should run. Photo: Jonn Leffmann, Wikimedia Commons, CC BY 4.0
November 2, 2026AI

AI Agent Sandboxes: Where Should Your Agent Run? (2026)

Where should an AI agent run code? A vendor-neutral guide to agent sandboxes: the threat model, microVMs vs containers, ...

View All Articles