[{"data":1,"prerenderedAt":1627},["ShallowReactive",2],{"site-nav-content":3,"blog:/blog/what-is-an-ai-workstream":163,"blog-index-copy":661,"blog:/blog/what-is-an-ai-workstream:surround":682,"hiring-banner-content":1596,"site-cta-content":1608},{"header":4,"productNav":9,"nav":42,"footer":61,"askAI":116,"id":147,"title":148,"archived":149,"authors":150,"badge":150,"body":151,"date":150,"department":150,"description":155,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":150,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":159,"relatedHeading":150,"seo":160,"series":150,"sitemap":115,"status":150,"stem":161,"subhead":150,"tags":150,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":162},{"productLabel":5,"loginLabel":6,"contactLabel":7,"contactSalesLabel":8},"Product","Log in","Contact","Get started for free",[10,14,18,22,26,30,34,38],{"label":11,"to":12,"description":13},"Overview","/overview","Seven layers. One closed loop.",{"label":15,"to":16,"description":17},"Conflux","/product/conflux","Where your team, workstreams, and agents meet.",{"label":19,"to":20,"description":21},"Agent Teams","/product/agent-teams","Specialist teams - governed from day one.",{"label":23,"to":24,"description":25},"Lifecycle Graph","/product/lifecycle-graph","Intelligence that compounds across every interaction.",{"label":27,"to":28,"description":29},"Company Wiki","/product/wiki","Playbooks and policies where expertise stays.",{"label":31,"to":32,"description":33},"Workstreams","/product/workstreams","From brief to signed-off deliverable on one canvas.",{"label":35,"to":36,"description":37},"Perception Console","/product/perception","Ask your whole business in plain English.",{"label":39,"to":40,"description":41},"Governance","/product/governance","Frontier AI you can actually sign off on.",[43,46,49,52,55,58],{"label":44,"to":45},"Models","/models",{"label":47,"to":48},"Pricing","/pricing",{"label":50,"to":51},"Integrations","/integrations",{"label":53,"to":54},"Security","/security",{"label":56,"to":57},"Partners","/partners",{"label":59,"to":60},"Insights","/blog",{"productHeading":5,"companyHeading":62,"legalHeading":63,"docsLabel":64,"docsUrl":65,"statementLines":66,"copyright":69,"companyLinks":70,"legalLinks":85,"socialLinks":95,"bottomLinks":105},"Company","Legal","Docs","https://docs.gonimbus.ai",[67,68],"Stop training someone else's model.","Control your AI.","© 2026 Nimbus Intelligence, Inc. All rights reserved.",[71,72,73,74,75,77,80,83],{"label":47,"to":48},{"label":50,"to":51},{"label":53,"to":54},{"label":59,"to":60},{"label":76,"to":57},"Partner Program",{"label":78,"to":79},"Careers","/careers",{"label":81,"to":82},"System status","/status",{"label":7,"to":84},"/contact",[86,89,92],{"label":87,"to":88},"Terms of Service","/terms",{"label":90,"to":91},"Privacy Policy","/privacy",{"label":93,"to":94},"Compliance","/compliance",[96,99,102],{"label":97,"href":98},"LinkedIn","https://www.linkedin.com/company/gonimbusai/",{"label":100,"href":101},"X","https://x.com/gonimbusai",{"label":103,"href":104},"Instagram","https://www.instagram.com/gonimbus_ai/",[106,108,110,111,112],{"label":107,"to":88},"Terms",{"label":109,"to":91},"Privacy",{"label":93,"to":94},{"label":81,"to":82},{"label":113,"to":114,"external":115},"LLMs.txt","/llms.txt",true,{"text":117,"prompt":118},"Ask AI about Nimbus",{"I'm researching enterprise intelligence platforms and want to know how Nimbus combines perception, collaboration, and autonomous agents to drive strategic decision-making":119,"platforms":121},{" Summarize the highlights from Nimbus's website":120},"https://gonimbus.ai",[122,127,132,137,142],{"name":123,"label":124,"icon":125,"hrefPrefix":126},"chatgpt","ChatGPT","simple-icons:openai","https://chatgpt.com/?prompt=",{"name":128,"label":129,"icon":130,"hrefPrefix":131},"perplexity","Perplexity","mdi:magnify","https://www.perplexity.ai/search/new?q=",{"name":133,"label":134,"icon":135,"hrefPrefix":136},"grok","Grok","simple-icons:x","https://x.com/i/grok?text=",{"name":138,"label":139,"icon":140,"hrefPrefix":141},"claude","Claude","simple-icons:anthropic","https://claude.ai/new?q=",{"name":143,"label":144,"icon":145,"hrefPrefix":146},"google-ai","Google AI","simple-icons:google","https://www.google.com/search?udm=50&aep=11&q=","content/shared/nav.md","Site navigation",false,null,{"type":152,"value":153,"toc":154},"minimark",[],{"title":155,"searchDepth":156,"depth":156,"links":157},"",2,[],"md","/shared/nav",{"title":148,"description":155},"shared/nav","p6IsjEfjcrwkjwvGsPcuEkpoJSeUn2vJXPTxLWQfw9M",{"id":164,"title":165,"archived":149,"authors":166,"badge":169,"body":171,"date":650,"department":150,"description":651,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":150,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":652,"relatedHeading":150,"seo":653,"series":654,"sitemap":115,"status":150,"stem":655,"subhead":150,"tags":656,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":660},"content/blog/what-is-an-ai-workstream.md","What is an AI Workstream",[167],{"name":168,"to":120},"Nimbus Research",{"label":170},"Explainer",{"type":152,"value":172,"toc":625},[173,182,185,188,199,202,207,293,301,305,311,314,333,336,361,364,369,375,384,390,396,405,409,415,420,426,432,438,444,447,455,459,466,469,476,491,495,499,502,506,509,513,516,520,526,530,535,539,542,546,553,557,560,564,571,575,578,582,585,589,594,598,607,611],[174,175,176,177,181],"p",{},"An AI workstream is a ",[178,179,180],"strong",{},"shared workspace for one job",": a brief, the tools allowed, the people and AI on it, a budget, and a finish line.",[174,183,184],{},"A Slack channel with a bot is not a job. It is a room. Anyone can paste anything. The bot never knows when the work is done. Next quarter, nobody can say which systems were in play or who was allowed to change them.",[174,186,187],{},"If you cannot name the systems in scope and the approval policy on writes, you do not have a workstream. You have a conversation.",[174,189,190,191,198],{},"Software teams already learned this. Work lives in issues and tickets, not in unbounded chat. ",[192,193,197],"a",{"href":194,"rel":195},"https://www.atlassian.com/agile/project-management/epics-stories-themes",[196],"nofollow","Atlassian’s epics and stories"," are named packages with a boundary. AI operations are still catching up. The missing object is often the work package: a place where the job actually lives.",[174,200,201],{},"The analogy is not decoration. Tickets have a requester, a scope, an owner, and a closed state. Copilots have a thread. Threads do not archive cleanly, do not attach least-privilege connectors, and do not carry a named signer. When AI started touching live systems, the thread stopped being a sufficient container.",[203,204,206],"h2",{"id":205},"words-youll-hear","Words you’ll hear",[208,209,210,217,223,240,246,257,263,273,283],"ul",{},[211,212,213,216],"li",{},[178,214,215],{},"Brief."," What this job is for, and what “done” means. At work, “Q3 regional discount hygiene” is a brief. “My stuff” is not.",[211,218,219,222],{},[178,220,221],{},"Connector."," A secure link to a live system (CRM, ERP, Drive). Attach what this job needs — not every system “just in case.” Default is read-only.",[211,224,225,228,229,233,234,239],{},[178,226,227],{},"Scope / least privilege."," Only the data and tools required for ",[230,231,232],"em",{},"this"," job. ",[192,235,238],{"href":236,"rel":237},"https://www.hhs.gov/hipaa/for-professionals/privacy/guidance/minimum-necessary-requirement/index.html",[196],"HIPAA’s minimum necessary"," is the same instinct: do not attach every system to every task. GDPR purpose limitation is the privacy-law cousin.",[211,241,242,245],{},[178,243,244],{},"God workspace."," One org-wide chat that can see every folder and every CRM object because setup was easier. At work, this is how recruiting sees finance forecasts.",[211,247,248,251,252,256],{},[178,249,250],{},"Agent team."," The AI specialists assigned to the job. The workstream is the stage; the team is the cast. See ",[192,253,255],{"href":254},"what-is-multi-agent-ai","What is multi-agent AI",".",[211,258,259,262],{},[178,260,261],{},"Named signer."," Who must approve a write. At work, this is a role that already owns that class of change.",[211,264,265,268,269,256],{},[178,266,267],{},"NTU / budget."," The spend ceiling for the job. See ",[192,270,272],{"href":271},"what-is-ai-token-economics","What is AI token economics",[211,274,275,278,279,256],{},[178,276,277],{},"Wiki section."," The asserted playbooks this job may load. See ",[192,280,282],{"href":281},"what-is-a-company-wiki-for-ai-agents","What is a company wiki for AI agents",[211,284,285,288,289,256],{},[178,286,287],{},"Lifecycle Graph."," The chain this job emits as it runs. See ",[192,290,292],{"href":291},"what-is-a-lifecycle-graph","What is a lifecycle graph",[174,294,295,296,300],{},"An ",[192,297,299],{"href":298},"what-is-an-agentic-workflow","agentic workflow"," is the sequence of steps. The workstream is the durable instance those steps run inside. A workflow definition without a workstream is a script on someone’s laptop. A workstream without a workflow is a folder with no process.",[203,302,304],{"id":303},"why-you-should-care","Why you should care",[174,306,307,308,310],{},"Without a boundary, two departments sharing an AI tool will either over-share (the recruiting job can see finance forecasts) or under-share (people export spreadsheets to personal ChatGPT). The workstream is the compromise: enough context to do ",[230,309,232],{}," job, not the whole company.",[174,312,313],{},"It affects you if work:",[208,315,316,319,322,325,328],{},[211,317,318],{},"touches more than one system",[211,320,321],{},"involves more than one role",[211,323,324],{},"can change a live record",[211,326,327],{},"needs a budget you can attribute",[211,329,330,331],{},"must still be explainable after people leave — see ",[192,332,292],{"href":291},[174,334,335],{},"Open workstreams the way you would open a ticket:",[208,337,338,345,352,355,358],{},[211,339,340,341,344],{},"One workstream per ",[178,342,343],{},"outcome",", not per person. “Q3 regional discount hygiene” can have several humans. “My stuff” cannot be governed or archived.",[211,346,347,348,351],{},"Attach the ",[178,349,350],{},"minimum"," connectors.",[211,353,354],{},"Set the write policy on day one, even if you start read-only.",[211,356,357],{},"Reuse templates, not last month’s chat thread.",[211,359,360],{},"Close or archive when the job ends. A sprint that never ends is not a sprint.",[174,362,363],{},"A standing “Ask AI” workstream with org-wide connectors recreates the copilot, including the blast radius.",[365,366,368],"h3",{"id":367},"what-changes-by-role","What changes by role",[174,370,371,374],{},[178,372,373],{},"Finance."," Chargeback becomes possible because the job is named. Close workstreams can attach ERP read-only, load the close checklist from the wiki, and keep GTM out of the ledger. A company-wide AI pool with no workstream attribution is a shared kitchen.",[174,376,377,380,381,383],{},[178,378,379],{},"Legal."," Scope is a processing purpose. A workstream for a renewal can include legal and go-to-market on ",[230,382,232],{}," goal without merging their entire universes. Legal also gets a closed state: when the job ends, retention follows the type of record instead of an immortal channel.",[174,385,386,389],{},[178,387,388],{},"Operations."," This is the ticket analogue they already wanted. Ops should refuse god workspaces, insist on a finish line, and treat human wait as a status, not a side conversation in Slack.",[174,391,392,395],{},[178,393,394],{},"Go-to-market."," Cross-functional launches finally have a place that is not a merged Slack. GTM still should not get finance’s ERP “for context.” Templates beat copying last quarter’s thread, which silently copies last quarter’s over-attached connectors.",[174,397,398,401,402,404],{},[178,399,400],{},"Security."," Least privilege is now a product object, not a memo. Connectors default to read-only. Adding a write path is a deliberate change to ",[230,403,232],{}," job, not a tenant-wide toggle. A workstream that never closes is a standing access grant.",[365,406,408],{"id":407},"what-people-get-wrong","What people get wrong",[174,410,411,414],{},[178,412,413],{},"One workstream per person."," You cannot archive “my stuff.” You cannot attribute it. You cannot apply least privilege.",[174,416,417,419],{},[178,418,244],{}," Setup is easier. Blast radius is the company.",[174,421,422,425],{},[178,423,424],{},"ChatGPT Project as the unit."," Some files, some instructions. Typically no connector-level least privilege, quoted writes, spend caps, or lasting record. Fine for personal research. Not an operations unit.",[174,427,428,431],{},[178,429,430],{},"Too small."," If setup exceeds the job, use a lighter sanctioned copilot path. Do not open a workstream to rewrite one sentence.",[174,433,434,437],{},[178,435,436],{},"Too large."," If you cannot explain the purpose in one sentence, or you keep attaching “one more connector,” split.",[174,439,440,443],{},[178,441,442],{},"Never closing."," Standing rooms recreate Slack, including the archaeology problem.",[174,445,446],{},"Good looks like: one outcome, minimum connectors, write policy on day one, wiki sections subscribed, budget capped, named signer, archive when done. Failure looks like an org-wide copilot with every OAuth grant and a channel that outlives the campaign.",[174,448,449,450,454],{},"The ",[192,451,453],{"href":452},"what-is-an-enterprise-ai-operating-system","enterprise AI OS"," metaphor is isolation plus I/O plus state. The workstream is the isolation unit. Without it, connectors, wiki, and agent teams have nowhere to attach that an auditor could name.",[203,456,458],{"id":457},"how-this-shows-up-in-nimbus","How this shows up in Nimbus",[174,460,461,462,465],{},"In Nimbus, workstreams are how ",[192,463,464],{"href":298},"agentic workflows"," become company objects rather than a file only one engineer can run.",[174,467,468],{},"Each workstream carries a brief, wiki sections (approved playbooks), connector attachments (read-only by default), agent team assignment, spend budget, release policy on writes, and nodes on the Lifecycle Graph.",[174,470,471,472,475],{},"Cross-department work is multiple teams on one workstream, not a merged Slack. Operators open this themselves; the point of an ",[192,473,474],{"href":452},"enterprise AI operating system"," is that the job folder is a product, not a forward-deployed spreadsheet.",[174,477,478,479,481,482,485,486,485,489,256],{},"See ",[192,480,31],{"href":32},". Related product: ",[192,483,484],{"href":20},"Agent teams",", ",[192,487,488],{"href":28},"Wiki",[192,490,39],{"href":40},[203,492,494],{"id":493},"questions-people-actually-ask","Questions people actually ask",[365,496,498],{"id":497},"is-a-chatgpt-project-a-workstream","Is a ChatGPT “Project” a workstream?",[174,500,501],{},"It is a weak analogue: some files, some custom instructions. It typically lacks connector-level least privilege, quoted writes, spend caps, and a lasting record. Useful for personal research. Not an operations unit.",[365,503,505],{"id":504},"how-small-is-too-small","How small is too small?",[174,507,508],{},"If the setup cost exceeds the job, use a lighter sanctioned copilot path. Do not create a workstream to rewrite one sentence.",[365,510,512],{"id":511},"how-large-is-too-large","How large is too large?",[174,514,515],{},"If you cannot explain the purpose in one sentence, or you keep attaching “one more connector,” split.",[365,517,519],{"id":518},"can-one-workstream-serve-multiple-departments","Can one workstream serve multiple departments?",[174,521,522,523,525],{},"Yes — go-to-market and legal on a renewal, for example. They share ",[230,524,232],{}," goal’s scope, not each other’s entire universe.",[365,527,529],{"id":528},"how-do-we-budget-them","How do we budget them?",[174,531,532,533,256],{},"Caps per workstream, plus an organisation pool. Chargeback by workstream beats “the AI bill.” See ",[192,534,272],{"href":271},[365,536,538],{"id":537},"is-a-slack-channel-with-a-bot-enough-if-we-add-a-approve-command","Is a Slack channel with a bot enough if we add a /approve command?",[174,540,541],{},"No. A command is not connector least privilege, a quoted payload, a durable chain, or an archive policy. It is still a room.",[365,543,545],{"id":544},"who-is-allowed-to-open-a-workstream","Who is allowed to open a workstream?",[174,547,548,549,256],{},"Whoever is allowed to open that class of job in analogue life — with the same instinct as who may open a ticket or a change request. An “AI team” bottleneck recreates the waitlist that causes ",[192,550,552],{"href":551},"what-is-shadow-ai","shadow AI",[365,554,556],{"id":555},"what-happens-when-the-job-ends","What happens when the job ends?",[174,558,559],{},"Close or archive. Revoke standing connector usefulness. Keep the reconstructable chain according to retention, not the entire chat.",[365,561,563],{"id":562},"do-we-need-a-workstream-for-read-only-analysis","Do we need a workstream for read-only analysis?",[174,565,566,567,256],{},"When the analysis crosses systems, roles, or must be replayed later, yes. When it is personal drafting with no live-system scope, a sanctioned copilot may be enough. See ",[192,568,570],{"href":569},"how-to-choose-between-a-copilot-and-a-work-os","How to choose between a copilot and a work OS",[365,572,574],{"id":573},"how-do-wiki-and-connectors-differ-inside-a-workstream","How do wiki and connectors differ inside a workstream?",[174,576,577],{},"Wiki is asserted policy the job must follow. Connectors are live systems the job may read (and, if enabled, write). Mixing them into one “knowledge” pile is how Drive folklore overwrites the playbook.",[365,579,581],{"id":580},"can-we-keep-one-standing-workstream-for-ask-anything","Can we keep one standing workstream for “ask anything”?",[174,583,584],{},"You can. You will recreate the copilot, including over-share. Standing Q&A belongs on a tightly scoped, read-only path if it exists at all.",[365,586,588],{"id":587},"how-does-this-relate-to-agent-teams","How does this relate to agent teams?",[174,590,591,592,256],{},"The workstream is the job. The agent team is the cast assigned to it. Changing the cast does not change the brief, the connectors, or the signer. See ",[192,593,255],{"href":254},[203,595,597],{"id":596},"related-reading","Related reading",[174,599,600,603,604,256],{},[192,601,602],{"href":452},"What is an enterprise AI operating system"," and ",[192,605,606],{"href":298},"What is an agentic workflow",[203,608,610],{"id":609},"sources","Sources",[208,612,613,619],{},[211,614,615],{},[192,616,618],{"href":194,"rel":617},[196],"Atlassian, epics, stories, and themes",[211,620,621],{},[192,622,624],{"href":236,"rel":623},[196],"HHS, HIPAA minimum necessary requirement",{"title":155,"searchDepth":156,"depth":156,"links":626},[627,628,633,634,648,649],{"id":205,"depth":156,"text":206},{"id":303,"depth":156,"text":304,"children":629},[630,632],{"id":367,"depth":631,"text":368},3,{"id":407,"depth":631,"text":408},{"id":457,"depth":156,"text":458},{"id":493,"depth":156,"text":494,"children":635},[636,637,638,639,640,641,642,643,644,645,646,647],{"id":497,"depth":631,"text":498},{"id":504,"depth":631,"text":505},{"id":511,"depth":631,"text":512},{"id":518,"depth":631,"text":519},{"id":528,"depth":631,"text":529},{"id":537,"depth":631,"text":538},{"id":544,"depth":631,"text":545},{"id":555,"depth":631,"text":556},{"id":562,"depth":631,"text":563},{"id":573,"depth":631,"text":574},{"id":580,"depth":631,"text":581},{"id":587,"depth":631,"text":588},{"id":596,"depth":156,"text":597},{"id":609,"depth":156,"text":610},"2026-08-17","An AI workstream is a shared workspace for one job: a brief, the tools allowed, the people and AI on it, a budget, and a finish line — not a Slack channel with a bot.","/blog/what-is-an-ai-workstream",{"title":165,"description":651},"explainer","blog/what-is-an-ai-workstream",[654,657,658,659],"workstreams","enterprise-ai","governance","Auw7pxXLBHz5skUcPAsGCeasYr2BpxrUjMrBsoRzRYo",{"hero":662,"id":664,"title":665,"archived":149,"authors":150,"badge":150,"body":666,"date":150,"department":150,"description":670,"extension":158,"eyebrow":671,"faqHeader":150,"faqs":150,"footerBand":672,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":60,"relatedHeading":678,"seo":679,"series":150,"sitemap":115,"status":150,"stem":680,"subhead":150,"tags":150,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":681},{"filename":663},"u2221455217_Flat_design_of_a_futuristic_minimalist_landscape__5d589295-cdea-4ea9-a262-be766881accf_1.png","content/blog/index.md","Exploring the future of intelligence.",{"type":152,"value":667,"toc":668},[],{"title":155,"searchDepth":156,"depth":156,"links":669},[],"Deep dives into pre-cognitive intelligence, sentient enterprises, and the evolving landscape of AI-driven business transformation.","Latest Research",{"headline":673,"description":674,"primaryLabel":675,"primaryTo":676,"secondaryLabel":677,"secondaryTo":12},"Stay at the frontier.","Subscribe for product updates and new insights.","Subscribe","/newsletter","Explore the platform","More research",{"title":665,"description":670},"blog/index","eK1RCXdDW8nfLSyKRXGB1mJm9FAmhAO6GWXwNKOMNVE",[683,1163],{"id":684,"title":685,"archived":149,"authors":686,"badge":688,"body":689,"date":650,"department":150,"description":693,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":150,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":1157,"relatedHeading":150,"seo":1158,"series":654,"sitemap":115,"status":150,"stem":1159,"subhead":150,"tags":1160,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":1162},"content/blog/what-is-an-enterprise-ai-operating-system.md","What is an Enterprise AI Operating System",[687],{"name":168,"to":120},{"label":170},{"type":152,"value":690,"toc":1133},[691,694,697,700,708,711,713,783,786,819,831,838,840,846,849,852,855,858,860,865,878,883,888,893,895,901,915,921,927,933,936,946,948,951,991,1006,1008,1012,1015,1019,1027,1031,1034,1038,1041,1045,1048,1052,1055,1059,1062,1066,1071,1075,1078,1082,1087,1091,1094,1098,1109,1111,1117,1119],[174,692,693],{},"An enterprise AI operating system is the layer between the AI model and how departments actually work — like Windows sits between the chip and your apps.",[174,695,696],{},"If your question is “which model should we buy,” you are shopping for a chip. If your question is “how do revenue, legal, and finance run the same loop without a personal-account workaround,” you are shopping for an OS.",[174,698,699],{},"It is not a chatbot with company login. It is not a model API with a prompt library.",[174,701,702,707],{},[192,703,706],{"href":704,"rel":705},"https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai",[196],"McKinsey’s 2025 State of AI survey"," found that 88% of companies use AI in at least one function — and that a majority are still piloting. About one in three report that they are scaling. Buying another model does not close that gap. The missing layer is how work actually runs.",[174,709,710],{},"The OS metaphor is useful if you keep it honest. An operating system does not replace your spreadsheet or your CRM. It gives applications isolation, permissions, input and output, and a place to keep state after the window closes. An enterprise AI OS does the same for work that uses models: isolation of jobs, rights over tools and data, reads and writes to live systems, budgets, and a record that survives the session.",[203,712,206],{"id":205},[208,714,715,721,727,737,743,755,765,775],{},[211,716,717,720],{},[178,718,719],{},"Copilot."," A high-quality assistant for a person. Admin controls, company login. Not, by itself, how several departments finish one job under a named signer. At work, this is “help me draft.” It is not “release this CRM change.”",[211,722,723,726],{},[178,724,725],{},"Operating system (in this sense)."," Process isolation, permissions, input/output to live systems, budgeting, and durable state — the jobs a kernel does for apps.",[211,728,729,732,733,736],{},[178,730,731],{},"System of record."," CRM, ERP, HR — still authoritative. The OS is the system of ",[230,734,735],{},"work",", not a second CRM.",[211,738,739,742],{},[178,740,741],{},"Forward-deployed engineer."," A vendor consultant who sits with you for months. Some programmes need that. Many companies need governed work this quarter without it.",[211,744,745,748,749,754],{},[178,746,747],{},"NIST AI RMF."," ",[192,750,753],{"href":751,"rel":752},"https://www.nist.gov/itl/ai-risk-management-framework",[196],"Govern, Map, Measure, Manage"," — public-sector language for the same kernel idea: identity, tool rights, and a record attached to real actions.",[211,756,757,760,761,256],{},[178,758,759],{},"Workstream."," The process-isolation unit: one job, one scope, one finish line. See ",[192,762,764],{"href":763},"what-is-an-ai-workstream","What is an AI workstream",[211,766,767,770,771,256],{},[178,768,769],{},"Fail-closed writes."," Missing named signer means nothing happens. See ",[192,772,774],{"href":773},"what-is-write-back-governance","What is write-back governance",[211,776,777,780,781,256],{},[178,778,779],{},"NTU."," A normalised work credit so spend can be quoted and capped. See ",[192,782,272],{"href":271},[174,784,785],{},"Five jobs cluster around the term:",[787,788,789,795,801,807,813],"ol",{},[211,790,791,794],{},[178,792,793],{},"Process isolation."," Go-to-market does not silently inherit finance’s ERP login.",[211,796,797,800],{},[178,798,799],{},"Resource management."," Inference and tool calls are budgeted.",[211,802,803,806],{},[178,804,805],{},"I/O control."," Reads and writes to CRM and ERP are first-class — not “chat that sometimes calls an API.”",[211,808,809,812],{},[178,810,811],{},"Permissioning."," Identity and context decide what an agent can see and do. A signed PDF is not enforcement.",[211,814,815,818],{},[178,816,817],{},"Durable state."," Outcomes, approvals, and rationale survive the session.",[174,820,821,822,603,826,830],{},"Copilots generally fail the last three. ",[192,823,825],{"href":824},"/blog/nimbus-vs-chatgpt-enterprise","ChatGPT Enterprise",[192,827,829],{"href":828},"/blog/nimbus-vs-claude","Claude for Work"," are excellent assistants. They are not this job.",[174,832,833,837],{},[192,834,836],{"href":835},"what-is-model-context-protocol","Model Context Protocol"," is also not this job. A common plug for tools is USB. USB did not create Windows.",[203,839,304],{"id":303},[174,841,842,843,845],{},"Operators do not “open the OS” the way they open a model playground. They open ",[178,844,735],{},": a brief, a scoped live system, a review, a release.",[174,847,848],{},"It affects you if AI is starting to touch revenue, financial close, customer records, or regulated processes. Chat history does not answer “who approved this, against which policy?”",[174,850,851],{},"The OS also matters if you refuse a six-to-twelve-month vendor-engineer programme as the only path to production.",[174,853,854],{},"The anti-pattern is using an OS as a better chatbot: one user, one thread, no write path, no memory beyond the conversation. If nobody except the original operator can reconstruct what happened, you have a log, not an operating system.",[174,856,857],{},"Personal copilots optimise for “the model always answers.” An OS optimises for “the company only acts when the gate says so.” A spend cap or a missing approval is a successful outcome.",[365,859,368],{"id":367},[174,861,862,864],{},[178,863,373],{}," The OS is how close and forecast jobs get a budget, a read-only ERP connector, a wiki checklist, and a named signer — without a second ledger. Finance should still own NetSuite. The OS should point at it.",[174,866,867,869,870,873,874,256],{},[178,868,379],{}," Reconstructable authorisation, purpose-limited scope, and a place that is not a personal chat vendor. Legal should evaluate whether unapproved writes are ",[230,871,872],{},"impossible",", not whether a policy PDF exists. See ",[192,875,877],{"href":876},"what-is-ai-governance","What is AI governance",[174,879,880,882],{},[178,881,388],{}," Isolation and durable state are ops problems. Ops should ask whether a paused run is a first-class object, whether connectors default to read-only, and whether Perception (or equivalent) can answer “why did this change?” without a data team reconstructing Slack.",[174,884,885,887],{},[178,886,394],{}," Cross-department loops — legal on a renewal, finance on a discount — need a shared job, not a shared inbox. GTM should not have to choose between a copilot that cannot write safely and a spreadsheet export to a consumer model.",[174,889,890,892],{},[178,891,400],{}," Identity, least privilege, fail-closed I/O, and not turning the OS into a second store of the whole company. Security also cares that self-service configuration does not mean tenant-wide write keys.",[365,894,408],{"id":407},[174,896,897,900],{},[178,898,899],{},"“ChatGPT with integrations.”"," Plugins without scoped work, approval architecture, and durable decision records are plugins. A copilot with automation actions can move data. It cannot, by itself, make unapproved writes impossible.",[174,902,903,906,907,910,911,914],{},[178,904,905],{},"MLOps as a substitute."," MLOps governs ",[230,908,909],{},"model production",". An enterprise AI OS governs ",[230,912,913],{},"operational work that uses models",". They stack.",[174,916,917,920],{},[178,918,919],{},"Replacing the CRM."," Salesforce, NetSuite, Workday, and the warehouse remain authoritative. Duplicating them is a second system of record.",[174,922,923,926],{},[178,924,925],{},"OS as chatbot."," One user, one thread, no write path, no memory. That is a copilot with extra vocabulary.",[174,928,929,932],{},[178,930,931],{},"Forward-deployed as the only path."," Some warehouses need specialists. Most operators need to attach a connector and set a named signer in the UI.",[174,934,935],{},"Good looks like: workstreams, wiki, read-only-default connectors, agent teams, Lifecycle Graph, model routing, NTU quotes, fail-closed writes, self-service configuration. Failure looks like another model contract plus a six-month SOW.",[174,937,938,939,941,942,256],{},"For the copilot-versus-OS choice, see ",[192,940,570],{"href":569},". For vendor scoring, ",[192,943,945],{"href":944},"how-to-evaluate-an-enterprise-ai-operating-system","How to evaluate an enterprise AI operating system",[203,947,458],{"id":457},[174,949,950],{},"Nimbus is a self-service enterprise AI OS. Operators configure it in the product.",[208,952,953,960,965,971,976,985],{},[211,954,955,959],{},[178,956,957],{},[192,958,31],{"href":763}," isolate process.",[211,961,962,964],{},[178,963,488],{}," holds asserted policy — approved playbooks, not a dump of PDFs a search might find.",[211,966,967,970],{},[178,968,969],{},"Connectors"," attach live systems. Default is read-only. Write-back is opt-in and gated.",[211,972,973,975],{},[178,974,484],{}," are department-shaped.",[211,977,978,980,981,984],{},[178,979,23],{}," stores the causal record. ",[178,982,983],{},"Perception"," queries it in ordinary language.",[211,986,987,990],{},[178,988,989],{},"Model routing"," puts routine extract on cheaper models.",[174,992,478,993,603,995,997,998,485,1000,485,1002,485,1004,256],{},[192,994,11],{"href":12},[192,996,945],{"href":944},". Product surfaces: ",[192,999,31],{"href":32},[192,1001,39],{"href":40},[192,1003,23],{"href":24},[192,1005,983],{"href":36},[203,1007,494],{"id":493},[365,1009,1011],{"id":1010},"is-an-enterprise-ai-os-just-chatgpt-with-integrations","Is an enterprise AI OS just “ChatGPT with integrations”?",[174,1013,1014],{},"No. Integrations without scoped work, approval architecture, and durable decision records are plugins. A copilot with automation actions can move data. It cannot, by itself, make unapproved writes impossible.",[365,1016,1018],{"id":1017},"how-is-this-different-from-mlops","How is this different from MLOps?",[174,1020,1021,1022,910,1024,1026],{},"MLOps governs ",[230,1023,909],{},[230,1025,913],{},". They stack. They do not substitute.",[365,1028,1030],{"id":1029},"do-we-still-need-a-crm-if-we-buy-an-os","Do we still need a CRM if we buy an OS?",[174,1032,1033],{},"Yes. Salesforce, NetSuite, Workday, and the warehouse remain authoritative.",[365,1035,1037],{"id":1036},"does-every-company-need-an-os","Does every company need an OS?",[174,1039,1040],{},"If the job is personal drafting with no writes to live systems, a governed copilot may be enough. The OS becomes the right abstraction when work crosses departments, when writes are material, and when you must reconstruct decisions.",[365,1042,1044],{"id":1043},"is-this-the-same-as-an-integration-platform-ipaas","Is this the same as an integration platform (iPaaS)?",[174,1046,1047],{},"No. iPaaS moves data on schedules and triggers. An AI OS runs language-using jobs with scope, spend, and a human gate. You may still need iPaaS. It does not quote a named signer on a CRM payload.",[365,1049,1051],{"id":1050},"does-operating-system-mean-we-install-software-on-laptops","Does “operating system” mean we install software on laptops?",[174,1053,1054],{},"No. It is a layer for work, not a desktop kernel. The metaphor is isolation, permissions, I/O, and state.",[365,1056,1058],{"id":1057},"can-we-build-this-ourselves-on-a-model-api","Can we build this ourselves on a model API?",[174,1060,1061],{},"You can assemble pieces. You will still need isolation, connectors, gates, spend, and a graph. Most “we built a GPT” programmes stall at the copilot layer. McKinsey’s split between using AI and scaling it is that stall in survey form.",[365,1063,1065],{"id":1064},"where-do-agent-teams-fit","Where do agent teams fit?",[174,1067,1068,1069,256],{},"They are the department-shaped specialists the OS schedules onto workstreams. They are not the OS. See ",[192,1070,255],{"href":254},[365,1072,1074],{"id":1073},"how-does-nists-ai-rmf-map","How does NIST’s AI RMF map?",[174,1076,1077],{},"Govern (owners, policy), Map (inventory of jobs and systems), Measure (evidence, spend, rejects), Manage (fail-closed writes, incident path). A product can make those cheaper. A framework PDF cannot enforce them.",[365,1079,1081],{"id":1080},"what-is-perception-in-this-picture","What is Perception in this picture?",[174,1083,1084,1085,256],{},"Ordinary-language questions over the company’s graph, wiki, and scoped systems — with the next step being a workstream, not another search. See ",[192,1086,983],{"href":36},[365,1088,1090],{"id":1089},"do-we-need-a-forward-deployed-engineer-to-go-live","Do we need a forward-deployed engineer to go live?",[174,1092,1093],{},"Not as the default path. If operators cannot attach a read-only connector and set a named signer in the UI, you do not have a self-service OS. Specialists belong on genuine exceptions, such as a warehouse with no OAuth.",[365,1095,1097],{"id":1096},"is-search-rag-an-os","Is search (RAG) an OS?",[174,1099,1100,1101,603,1105,256],{},"No. Lookup-then-answer is infrastructure. It does not isolate jobs or gate writes. See ",[192,1102,1104],{"href":1103},"what-is-enterprise-rag","What is enterprise RAG",[192,1106,1108],{"href":1107},"/blog/nimbus-vs-glean","Nimbus vs Glean",[203,1110,597],{"id":596},[174,1112,1113,603,1115,256],{},[192,1114,764],{"href":763},[192,1116,877],{"href":876},[203,1118,610],{"id":609},[208,1120,1121,1127],{},[211,1122,1123],{},[192,1124,1126],{"href":704,"rel":1125},[196],"McKinsey, The state of AI (2025)",[211,1128,1129],{},[192,1130,1132],{"href":751,"rel":1131},[196],"NIST AI Risk Management Framework",{"title":155,"searchDepth":156,"depth":156,"links":1134},[1135,1136,1140,1141,1155,1156],{"id":205,"depth":156,"text":206},{"id":303,"depth":156,"text":304,"children":1137},[1138,1139],{"id":367,"depth":631,"text":368},{"id":407,"depth":631,"text":408},{"id":457,"depth":156,"text":458},{"id":493,"depth":156,"text":494,"children":1142},[1143,1144,1145,1146,1147,1148,1149,1150,1151,1152,1153,1154],{"id":1010,"depth":631,"text":1011},{"id":1017,"depth":631,"text":1018},{"id":1029,"depth":631,"text":1030},{"id":1036,"depth":631,"text":1037},{"id":1043,"depth":631,"text":1044},{"id":1050,"depth":631,"text":1051},{"id":1057,"depth":631,"text":1058},{"id":1064,"depth":631,"text":1065},{"id":1073,"depth":631,"text":1074},{"id":1080,"depth":631,"text":1081},{"id":1089,"depth":631,"text":1090},{"id":1096,"depth":631,"text":1097},{"id":596,"depth":156,"text":597},{"id":609,"depth":156,"text":610},"/blog/what-is-an-enterprise-ai-operating-system",{"title":685,"description":693},"blog/what-is-an-enterprise-ai-operating-system",[654,658,1161,659],"operating-system","KvG0Zq4odG5KLc9Qvg-KiIab5qpFy43zV4peSOw3woY",{"id":1164,"title":1165,"archived":149,"authors":1166,"badge":1168,"body":1169,"date":650,"department":150,"description":1588,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":150,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":1589,"relatedHeading":150,"seo":1590,"series":654,"sitemap":115,"status":150,"stem":1591,"subhead":150,"tags":1592,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":1595},"content/blog/what-is-an-agentic-workflow.md","What is an Agentic Workflow",[1167],{"name":168,"to":120},{"label":170},{"type":152,"value":1170,"toc":1564},[1171,1174,1184,1192,1195,1197,1200,1222,1225,1272,1282,1284,1287,1294,1297,1331,1334,1336,1341,1352,1357,1362,1377,1379,1385,1391,1397,1403,1411,1414,1416,1423,1430,1439,1441,1445,1448,1452,1455,1459,1462,1466,1469,1473,1478,1482,1485,1489,1492,1496,1503,1507,1510,1514,1519,1523,1528,1532,1535,1537,1543,1545],[174,1172,1173],{},"“We have an agent” often means a chat that never knows when to stop. Someone types a goal. The model keeps calling tools until the budget dies, or until a human closes the tab. There is no finish line. There is a conversation that looked busy.",[174,1175,1176,1177,1180,1181,1183],{},"A ",[178,1178,1179],{},"workflow"," has steps and a stop. An ",[178,1182,299],{}," is a sequence of steps an AI can run toward a goal, with rules for when to stop — including a person who must approve before a live system changes.",[174,1185,1186,1191],{},[192,1187,1190],{"href":1188,"rel":1189},"https://www.anthropic.com/engineering/building-effective-agents",[196],"Anthropic’s note on building effective agents"," makes the same cut: workflows with tools and stop conditions, not endless chat. The note is worth reading because it is honest about the boring parts — encoding the job, bounding the tools, and deciding what “done” means — rather than treating fluency as a process.",[174,1193,1194],{},"Older automation without models is brittle but auditable. Models without a workflow are flexible but unaccountable. An agentic workflow is the attempt to get both: language where the input is messy, and a finish line where the company needs one.",[203,1196,206],{"id":205},[174,1198,1199],{},"Vendors collapse three different layers into the word “agentic”:",[208,1201,1202,1208,1214],{},[211,1203,1204,1207],{},[178,1205,1206],{},"Agentic capability."," The model can use tools, plan, and reflect. At work, this is “it can search Drive and draft a note.” It is not yet a job.",[211,1209,1210,1213],{},[178,1211,1212],{},"Agentic workflow."," A designed sequence of those capabilities, with business stop conditions. This article is about this layer. At work, this is “extract, compare to the playbook, quote the CRM fields, wait for the named signer, write or refuse.”",[211,1215,1216,1219,1220,256],{},[178,1217,1218],{},"Agent platform."," Identity, connectors, tests, and governance around many workflows. At work, this is closer to an ",[192,1221,474],{"href":452},[174,1223,1224],{},"Other terms:",[208,1226,1227,1233,1239,1247,1257,1263],{},[211,1228,1229,1232],{},[178,1230,1231],{},"Tool."," An action the AI can take: search files, query CRM, post a message. At work, a tool is a hand. Hands are not roles, and they are not stop conditions.",[211,1234,1235,1238],{},[178,1236,1237],{},"Stop condition."," Budget hit, waiting on approval, error, empty result, human cancel. “The model says it is done” is a weak stop by itself.",[211,1240,1241,1244,1245,256],{},[178,1242,1243],{},"Write-back."," The AI is allowed to change a live system, not just draft. See ",[192,1246,774],{"href":773},[211,1248,1249,1252,1253,256],{},[178,1250,1251],{},"Human wait."," A step in the sequence, not an interruption. See ",[192,1254,1256],{"href":1255},"what-is-human-in-the-loop-ai","What is human-in-the-loop AI",[211,1258,1259,1262],{},[178,1260,1261],{},"Version."," Which workflow definition ran. When policy changes, retrieval changes. Operators need to know which version ran last Tuesday.",[211,1264,1265,1268,1269,256],{},[178,1266,1267],{},"MCP."," A common plug so AI apps can use the same tools. Plumbing. It does not define your stops. See ",[192,1270,1271],{"href":835},"What is Model Context Protocol",[174,1273,1176,1274,1277,1278,1281],{},[192,1275,1276],{"href":763},"workstream"," is the company object that ",[230,1279,1280],{},"hosts"," the workflow: brief, connectors, people, budget, finish line. The workflow is the sequence. The workstream is the job folder. Mixing those two words is how demos skip isolation.",[203,1283,304],{"id":303},[174,1285,1286],{},"Capability demos look like workflows. They are not. A fluent plan is not a paused run waiting on approval, a failed run that did not retry a write, or a replay of which step ran.",[174,1288,1289,1290,1293],{},"It affects you if the job is ",[178,1291,1292],{},"multi-step, tool-using, and repeated"," — the opposite of one-off chat. Close checklists, renewal playbooks, and incident runbooks already have steps. Encode those. If the job is not written down, you will encode folklore and then fight the folklore.",[174,1295,1296],{},"Practical rules:",[208,1298,1299,1305,1311,1317,1323],{},[211,1300,1301,1304],{},[178,1302,1303],{},"Read-heavy workflows"," can be long. They should still finish in an artefact with sources.",[211,1306,1307,1310],{},[178,1308,1309],{},"Write-heavy workflows"," should be short after the quote: one payload, one gate, one execution, one record. Do not hide ten writes in a “cleanup agent.”",[211,1312,1313,1316],{},[178,1314,1315],{},"Human wait is a step",", not an interruption.",[211,1318,1319,1322],{},[178,1320,1321],{},"Version the workflow."," Policy and retrieval drift. Last Tuesday’s run needs a definition you can still open.",[211,1324,1325,1328,1329,256],{},[178,1326,1327],{},"Budget is a stop."," See ",[192,1330,272],{"href":271},[174,1332,1333],{},"A mega-agent with “figure it out” as the spec is not a workflow. It is a hope.",[365,1335,368],{"id":367},[174,1337,1338,1340],{},[178,1339,373],{}," Close and forecast jobs already have checklists. An agentic workflow that posts a journal without a stop at the named signer is not “agentic.” It is unattended posting. Finance also needs spend stops so a retry loop cannot become the month’s inference bill.",[174,1342,1343,1345,1346,1351],{},[178,1344,379],{}," Customer-facing steps and anything that asserts a term need a gate before send. Air Canada’s chatbot invented a bereavement fare and the company was held to it — ",[192,1347,1350],{"href":1348,"rel":1349},"https://www.cbc.ca/news/canada/british-columbia/air-canada-chatbot-lawsuit-1.7116416",[196],"CBC’s report"," is the cautionary case for “the workflow ended at the message.” Legal also cares that the workflow version is reconstructable.",[174,1353,1354,1356],{},[178,1355,388],{}," This is the native language: runbooks, queues, retries, and “do not proceed.” Ops should refuse workflows that cannot pause cleanly, cannot show which step failed, and cannot distinguish “waiting on a person” from “waiting on a tool error.”",[174,1358,1359,1361],{},[178,1360,394],{}," Renewal and hygiene jobs are repeated and tool-using. GTM should demand a short write path after the quote, not a weekend “cleanup” that touches hundreds of records behind one click. Time-to-approved-write is the metric, not time-to-first-plan.",[174,1363,1364,1366,1367,1370,1371,1376],{},[178,1365,400],{}," Tool belts are attack surface. Prompt injection that tricks a model into ",[230,1368,1369],{},"requesting"," a write should still die at a fail-closed gate. Importing every MCP helper into one workflow is how a demo becomes one actor with every production login. The ",[192,1372,1375],{"href":1373,"rel":1374},"https://genai.owasp.org/llm-top-10/",[196],"OWASP Top 10 for LLM applications"," treats tool use as a security topic for this reason.",[365,1378,408],{"id":407},[174,1380,1381,1384],{},[178,1382,1383],{},"Chat as workflow."," A conversation that looks busy has no durable instance, no version, and no gate.",[174,1386,1387,1390],{},[178,1388,1389],{},"A checklist in a prompt."," A start. Without tools, a durable job, and a stop, it is still a prompt.",[174,1392,1393,1396],{},[178,1394,1395],{},"Replacing a stable bot."," If the job is a scheduled export, older automation is the right tool. Agentic workflows help on messy documents. They are not a prestige upgrade for a cron job.",[174,1398,1399,1402],{},[178,1400,1401],{},"Fully autonomous production."," Only for actions you would already automate without a model, plus logging. If you would not let a scheduled job do it, do not let an agent do it unattended.",[174,1404,1405,1408,1409,256],{},[178,1406,1407],{},"Multi-agent as a requirement."," A single tool-using agent can execute a workflow. Multiple agents help when duties already split. See ",[192,1410,255],{"href":254},[174,1412,1413],{},"Good looks like: named steps, bounded tools, explicit stops (including human wait and budget), versioned definitions, read-only by default, fail-closed writes. Failure looks like a flagship model with every connector and a spec that says “be helpful.”",[203,1415,458],{"id":457},[174,1417,1418,1419,256],{},"Nimbus’s delivery unit for operators is the ",[178,1420,1421],{},[192,1422,1276],{"href":763},[174,1424,1425,1426,1429],{},"The mapping in everyday terms: the brief is the goal; ",[192,1427,1428],{"href":254},"agent teams"," run the steps; connectors are the tools (default read-only); wiki is the playbook the steps must respect; governance is the wait/write stop; the Lifecycle Graph is the executed run. Model routing chooses the brain per step; it does not choose the stop.",[174,1431,478,1432,485,1434,1436,1437,256],{},[192,1433,31],{"href":32},[192,1435,484],{"href":20},", and ",[192,1438,39],{"href":40},[203,1440,494],{"id":493},[365,1442,1444],{"id":1443},"is-a-checklist-in-a-prompt-an-agentic-workflow","Is a checklist in a prompt an agentic workflow?",[174,1446,1447],{},"It is a start. If there are no tools, no durable instance, and no gate, it is a prompt.",[365,1449,1451],{"id":1450},"how-is-this-different-from-older-robotic-automation","How is this different from older robotic automation?",[174,1453,1454],{},"Older automation executes deterministic steps. Agentic workflows add language and planning. That helps on messy documents. It also means you need tests and human gates. Do not replace a stable bot with an agent if the job is still a scheduled export.",[365,1456,1458],{"id":1457},"do-agentic-workflows-require-multiple-agents","Do agentic workflows require multiple agents?",[174,1460,1461],{},"No. A single tool-using agent can execute a workflow. Multiple agents help when duties already split in the organisation.",[365,1463,1465],{"id":1464},"can-a-workflow-be-fully-autonomous-in-production","Can a workflow be fully autonomous in production?",[174,1467,1468],{},"Only for actions you would already automate without a model, plus logging.",[365,1470,1472],{"id":1471},"where-do-tool-connection-standards-fit","Where do tool-connection standards fit?",[174,1474,1475,1476,256],{},"A common plug so AI apps can use the same tools is plumbing. It does not define your stops or approvals. See ",[192,1477,1271],{"href":835},[365,1479,1481],{"id":1480},"what-is-a-good-stop-condition-besides-the-model-is-done","What is a good stop condition besides “the model is done”?",[174,1483,1484],{},"Budget ceiling, empty retrieval, tool error, human cancel, and wait-for-named-signer. “Done” from the model is a suggestion. Encode the others.",[365,1486,1488],{"id":1487},"how-long-should-a-write-heavy-workflow-be","How long should a write-heavy workflow be?",[174,1490,1491],{},"Short after the quote. One payload, one gate, one execution, one record. Length belongs in the read and compare steps, not in a bundle of hidden mutations.",[365,1493,1495],{"id":1494},"how-do-we-version-a-workflow-when-the-wiki-changes","How do we version a workflow when the wiki changes?",[174,1497,1498,1499,1502],{},"Treat the playbook version as an input to the run. The ",[192,1500,1501],{"href":291},"lifecycle graph"," should cite which wiki version the steps respected. Changing policy without recording which definition ran is how Tuesday becomes unexplained.",[365,1504,1506],{"id":1505},"is-agentic-the-same-as-autonomous","Is “agentic” the same as “autonomous”?",[174,1508,1509],{},"No. Agentic means the model can plan and use tools. Autonomy is a policy about whether a person must still sign. Most production writes should not be autonomous.",[365,1511,1513],{"id":1512},"can-we-import-every-available-tool-and-let-the-model-choose","Can we import every available tool and let the model choose?",[174,1515,1516,1517,256],{},"That is a confused workflow. Least privilege applies to tools as much as to data. See ",[192,1518,764],{"href":763},[365,1520,1522],{"id":1521},"how-does-this-relate-to-human-in-the-loop","How does this relate to human-in-the-loop?",[174,1524,1525,1526,256],{},"Human wait is a first-class step. If the person is only “on the loop” with a kill switch, you have a different design. See ",[192,1527,1256],{"href":1255},[365,1529,1531],{"id":1530},"will-a-better-model-remove-the-need-for-a-workflow","Will a better model remove the need for a workflow?",[174,1533,1534],{},"Stronger models plan more fluently. They still do not know your finish line, your signer, or your budget. Fluency without stops is a more expensive loop.",[203,1536,597],{"id":596},[174,1538,1539,603,1541,256],{},[192,1540,602],{"href":452},[192,1542,255],{"href":254},[203,1544,610],{"id":609},[208,1546,1547,1553,1559],{},[211,1548,1549],{},[192,1550,1552],{"href":1188,"rel":1551},[196],"Anthropic, Building effective agents",[211,1554,1555],{},[192,1556,1558],{"href":1348,"rel":1557},[196],"CBC, Air Canada chatbot lawsuit",[211,1560,1561],{},[192,1562,1375],{"href":1373,"rel":1563},[196],{"title":155,"searchDepth":156,"depth":156,"links":1565},[1566,1567,1571,1572,1586,1587],{"id":205,"depth":156,"text":206},{"id":303,"depth":156,"text":304,"children":1568},[1569,1570],{"id":367,"depth":631,"text":368},{"id":407,"depth":631,"text":408},{"id":457,"depth":156,"text":458},{"id":493,"depth":156,"text":494,"children":1573},[1574,1575,1576,1577,1578,1579,1580,1581,1582,1583,1584,1585],{"id":1443,"depth":631,"text":1444},{"id":1450,"depth":631,"text":1451},{"id":1457,"depth":631,"text":1458},{"id":1464,"depth":631,"text":1465},{"id":1471,"depth":631,"text":1472},{"id":1480,"depth":631,"text":1481},{"id":1487,"depth":631,"text":1488},{"id":1494,"depth":631,"text":1495},{"id":1505,"depth":631,"text":1506},{"id":1512,"depth":631,"text":1513},{"id":1521,"depth":631,"text":1522},{"id":1530,"depth":631,"text":1531},{"id":596,"depth":156,"text":597},{"id":609,"depth":156,"text":610},"An agentic workflow is a sequence of steps an AI can run toward a goal, with rules for when to stop — including a person who must approve before a live system changes.","/blog/what-is-an-agentic-workflow",{"title":1165,"description":1588},"blog/what-is-an-agentic-workflow",[654,1593,1594,657],"agentic-workflow","agents","OD070bDpNh56Kq9pajGR0tbzXSES_KnS9029g0lELm8",{"enabled":149,"message":1597,"linkLabel":78,"linkHref":79,"id":1598,"title":1599,"archived":149,"authors":150,"badge":150,"body":1600,"date":150,"department":150,"description":155,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":150,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":1604,"relatedHeading":150,"seo":1605,"series":150,"sitemap":115,"status":150,"stem":1606,"subhead":150,"tags":150,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":1607},"We're hiring! Join the team building the Sentient Enterprise.","content/shared/hiring.md","Hiring banner",{"type":152,"value":1601,"toc":1602},[],{"title":155,"searchDepth":156,"depth":156,"links":1603},[],"/shared/hiring",{"title":1599,"description":155},"shared/hiring","-6bioYD7lKYokGUVU3ff4hHTvB-sDyOMuCptKHnojfk",{"fold":1609,"id":1613,"title":1614,"archived":149,"authors":150,"badge":150,"body":1615,"date":150,"department":150,"description":155,"extension":158,"eyebrow":150,"faqHeader":150,"faqs":150,"footerBand":1619,"headline":150,"image":150,"industry":150,"jobType":150,"listed":149,"location":150,"navigation":115,"openRoles":150,"pageLayout":150,"path":1623,"relatedHeading":150,"seo":1624,"series":150,"sitemap":115,"status":150,"stem":1625,"subhead":150,"tags":150,"video":150,"whyJoin":150,"workplaceType":150,"__hash__":1626},{"headline":1610,"description":1611,"primaryLabel":8,"primaryTo":1612,"secondaryLabel":677,"secondaryTo":12},"Run frontier AI your business actually owns.","Governed agent swarms, 2,000+ integrations, and a knowledge graph that stays inside your walls. Free 7-day trial.","/checkout","content/shared/cta.md","Site CTAs",{"type":152,"value":1616,"toc":1617},[],{"title":155,"searchDepth":156,"depth":156,"links":1618},[],{"headline":1620,"description":1621,"primaryLabel":8,"primaryTo":1612,"secondaryLabel":1622,"secondaryTo":84},"See what governed AI looks like on your stack.","Connect your tools, run a workstream, and keep every decision on your ledger - free for 7 days.","Talk to our team","/shared/cta",{"title":1614,"description":155},"shared/cta","YHK6Fb8AvCPR1zZq7R_xiXUG0hwhP5UxHA8Ix52JQp4",1787194073125]