[{"data":1,"prerenderedAt":941},["ShallowReactive",2],{"insight-insights_en-the-30-day-playbook-week-by-week":3,"insight-related-insights_en-the-30-day-playbook-week-by-week":238},{"id":4,"title":5,"author":6,"blobHue":10,"body":11,"category":212,"date":213,"description":17,"draft":214,"extension":215,"featured":214,"headline":216,"hue":217,"letter":218,"meta":219,"navigation":220,"path":221,"readMinutes":222,"related":223,"seo":227,"stem":228,"summary":229,"toc":230,"__hash__":237},"insights_en\u002Finsights\u002Fthe-30-day-playbook-week-by-week.md","The 30-day playbook, week by week",{"name":7,"role":8,"bio":9},"André","Founder & CTO","André is the founder and CTO of WizardingCode. Eight years building the software companies run on, now putting agents into production.",null,{"type":12,"value":13,"toc":202},"minimark",[14,18,21,26,29,32,36,39,68,71,75,78,81,84,88,91,94,182,186,189,192,196,199],[15,16,17],"p",{},"Thirty days sounds fast for putting AI agents into production. It is fast. It is also the reason most of our projects ship: a fixed date forces every decision that a pilot would postpone to be taken in the first week.",[15,19,20],{},"This is the playbook we follow. Every step ends with something written down and someone’s name next to it.",[22,23,25],"h2",{"id":24},"week-0-diagnose","Week 0: diagnose",[15,27,28],{},"It starts with a free call. If there is a fit, we spend the next week inside your operations, most of it with the people who do the work. We sit in on the inbox, the report, the supplier file. We want to see where the hours go and where the money leaks, not how the process is described in a slide.",[15,30,31],{},"The artefact is a map of your operations with the value of every agent we would build, and a fixed price. The sign-off is yours: which process goes first. We push for one with volume, clear rules and a number someone already cares about.",[22,33,35],{"id":34},"week-1-map","Week 1: map",[15,37,38],{},"This is the week most pilots skip, and the week that decides whether yours ships.",[40,41,42,50,56,62],"ol",{},[43,44,45,49],"li",{},[46,47,48],"strong",{},"Access."," Real credentials to the real systems: the CRM, the helpdesk, the ERP, the inbox. Not an export.",[43,51,52,55],{},[46,53,54],{},"Real cases."," A set of past emails, tickets or files, with what a good answer looked like. They become the tests.",[43,57,58,61],{},[46,59,60],{},"One metric, signed off."," First response time, time to spot a new listing, partner hours on admin. One number, owned by the team whose work changes, signed by the person who owns it.",[43,63,64,67],{},[46,65,66],{},"First approval rules."," What the agents may do alone, and what always goes to a person, written in plain language.",[15,69,70],{},"If week 1 ends without access and a signed metric, we say so, because it is cheaper for both sides than finding out on day 29.",[22,72,74],{"id":73},"week-2-build","Week 2: build",[15,76,77],{},"Now we build the agents, the rules, the shared memory and the command center, wired into your tools. The agents live where your team already works; nobody gets a new app to learn.",[15,79,80],{},"The real cases from week 1 become evals: automated tests that run every answer against what a good answer looked like. Customer-facing answers get a judge. Tone is agreed with the people who own it, often in a single working session with a pile of past replies.",[15,82,83],{},"The sign-off at the end of the week is a walkthrough with the team lead, on their own cases.",[22,85,87],{"id":86},"week-3-prove","Week 3: prove",[15,89,90],{},"First the evals, on real cases the agents have never seen. Then shadow mode on live work: the agents draft, people send. Every edit a person makes is a signal, and we read them daily.",[15,92,93],{},"This is also where the handoff rules are tuned with the team. Which cases go straight to a person, at what threshold, through which channel. The team should be able to change a rule themselves, and in week 3 they practise doing it.",[95,96,97,113],"table",{},[98,99,100],"thead",{},[101,102,103,107,110],"tr",{},[104,105,106],"th",{},"STEP",[104,108,109],{},"ARTEFACT",[104,111,112],{},"WHO SIGNS",[114,115,116,130,143,156,169],"tbody",{},[101,117,118,124,127],{},[119,120,121],"td",{},[46,122,123],{},"Week 0",[119,125,126],{},"Operations map, value per agent, fixed price",[119,128,129],{},"You: which process first",[101,131,132,137,140],{},[119,133,134],{},[46,135,136],{},"Week 1",[119,138,139],{},"Access, real cases, first approval rules",[119,141,142],{},"The owner of the metric",[101,144,145,150,153],{},[119,146,147],{},[46,148,149],{},"Week 2",[119,151,152],{},"Agents, memory, rules, command center, evals",[119,154,155],{},"The team lead, after a walkthrough",[101,157,158,163,166],{},[119,159,160],{},[46,161,162],{},"Week 3",[119,164,165],{},"Eval results, shadow-mode log, tuned handoffs",[119,167,168],{},"The team lead: ready to go live",[101,170,171,176,179],{},[119,172,173],{},[46,174,175],{},"Day 30",[119,177,178],{},"Production, monitoring, trained team",[119,180,181],{},"Both of us, against the metric",[22,183,185],{"id":184},"day-30-live","Day 30: live",[15,187,188],{},"On day 30 the agents are in production, on live work, monitored from the command center. The team is trained on it: how to read a run, how to approve, how to pause an agent, how to change a rule.",[15,190,191],{},"Production means the agents answer, route or update on their own inside the rules, and the number from week 1 is tracked every day from then on. It does not mean “available on request” or “ready for phase two”.",[22,193,195],{"id":194},"why-the-date-holds","Why the date holds",[15,197,198],{},"Three things keep the date honest. The scope is one process, not the company. The metric is agreed before a line of code, so nobody can move the goalposts later. And we put our money on it: live in 30 days, or your money back.",[15,200,201],{},"After day 30 we run the fleet: monitoring, tuning and, when you are ready, the next squad. On our largest project that became one new squad a month, each built on the memory and rules of the ones before it.",{"title":203,"searchDepth":204,"depth":204,"links":205},"",2,[206,207,208,209,210,211],{"id":24,"depth":204,"text":25},{"id":34,"depth":204,"text":35},{"id":73,"depth":204,"text":74},{"id":86,"depth":204,"text":87},{"id":184,"depth":204,"text":185},{"id":194,"depth":204,"text":195},"playbooks","2026-08-18",false,"md","The 30-day playbook, week by ==week.==","magenta","30",{},true,"\u002Finsights\u002Fthe-30-day-playbook-week-by-week",4,[224,225,226],"why-95-percent-of-ai-pilots-never-reach-production","what-agents-should-never-do-alone","build-buy-or-both",{"title":5,"description":17},"insights\u002Fthe-30-day-playbook-week-by-week","From the first call to production: the exact steps, artefacts and sign-offs we use.",[231,232,233,234,235,236],{"id":24,"label":25},{"id":34,"label":35},{"id":73,"label":74},{"id":86,"label":87},{"id":184,"label":185},{"id":194,"label":195},"rT6HBQ-x40ciVZH-qHpcDH1wLuGv79MQR2E3ffHF_dY",[239,455,695],{"id":240,"title":241,"author":242,"blobHue":243,"body":244,"category":433,"date":434,"description":435,"draft":214,"extension":215,"featured":220,"headline":436,"hue":217,"letter":437,"meta":438,"navigation":220,"path":439,"readMinutes":440,"related":441,"seo":445,"stem":446,"summary":447,"toc":448,"__hash__":454},"insights_en\u002Finsights\u002Fwhy-95-percent-of-ai-pilots-never-reach-production.md","Why 95% of AI pilots never reach production",{"name":7,"role":8,"bio":9},"blue",{"type":12,"value":245,"toc":426},[246,254,257,261,266,272,276,279,299,303,306,387,391,413,417,420],[15,247,248,249,253],{},"In 2025, MIT’s NANDA initiative looked at how companies were using generative AI and found that about 95% of enterprise pilots delivered no measurable impact on the P&L.",[250,251,252],"sup",{},"1"," Billions spent, demos applauded, and almost nothing changed in how the business actually ran.",[15,255,256],{},"We’ve seen the same pattern from the inside. Most of the companies that call us have already run a pilot. It worked in the demo. It never went live.",[22,258,260],{"id":259},"the-number-everyone-quotes","The number everyone quotes",[15,262,263,264],{},"The headline is easy to misread. It doesn’t say AI doesn’t work. It says pilots don’t turn into production. The same report found that projects built with specialised external partners reached deployment about twice as often as internal builds, and that the biggest returns came from unglamorous back-office work, not from customer-facing chatbots.",[250,265,252],{},[267,268,269],"blockquote",{},[15,270,271],{},"The model is rarely the problem. The plumbing is.",[22,273,275],{"id":274},"its-not-the-model","It’s not the model",[15,277,278],{},"When we look at why a pilot stalled, the answer is almost never “the AI wasn’t smart enough”. It is one of three things:",[40,280,281,287,293],{},[43,282,283,286],{},[46,284,285],{},"No live data."," The pilot ran on an export. Connecting it to the real ERP, inbox or helpdesk was “phase two”, and phase two never came.",[43,288,289,292],{},[46,290,291],{},"No owner."," The innovation team built it. The team whose work it would take never asked for it, and never adopted it.",[43,294,295,298],{},[46,296,297],{},"No number."," Success was “a good demo”. Nobody agreed what had to move, so nobody could say it had worked.",[22,300,302],{"id":301},"four-things-the-5-do","Four things the 5% do",[15,304,305],{},"The projects that ship look different from day one. They connect to live systems in week one, not month six. The metric is agreed before a line of code is written, and it belongs to the team whose work changes. People approve the decisions that carry risk, so nobody has to trust the agent blindly. And there is a date: production in weeks, not a roadmap.",[95,307,308,320],{},[98,309,310],{},[101,311,312,314,317],{},[104,313],{},[104,315,316],{},"A PILOT",[104,318,319],{},"A SYSTEM IN PRODUCTION",[114,321,322,335,348,361,374],{},[101,323,324,329,332],{},[119,325,326],{},[46,327,328],{},"Data",[119,330,331],{},"A sample export",[119,333,334],{},"Your live systems",[101,336,337,342,345],{},[119,338,339],{},[46,340,341],{},"Success",[119,343,344],{},"A good demo",[119,346,347],{},"A number agreed on day one",[101,349,350,355,358],{},[119,351,352],{},[46,353,354],{},"Owner",[119,356,357],{},"The innovation team",[119,359,360],{},"The team whose work it takes",[101,362,363,368,371],{},[119,364,365],{},[46,366,367],{},"Humans",[119,369,370],{},"Watching",[119,372,373],{},"Approving what matters",[101,375,376,381,384],{},[119,377,378],{},[46,379,380],{},"Timeline",[119,382,383],{},"Open-ended",[119,385,386],{},"Weeks",[22,388,390],{"id":389},"a-checklist-before-you-start","A checklist before you start",[392,393,395],"prose-checklist",{"title":394},"Before you approve the next AI project, ask",[396,397,398,401,404,407,410],"ul",{},[43,399,400],{},"Which live system does it read and write, in week one?",[43,402,403],{},"Which number moves, and who owns it?",[43,405,406],{},"Which decisions stay with a person?",[43,408,409],{},"What happens on day 30?",[43,411,412],{},"Who owns the code, the prompts and the data?",[22,414,416],{"id":415},"what-this-means-for-you","What this means for you",[15,418,419],{},"If your last pilot is gathering dust, it probably wasn’t a bad idea. It was missing the boring parts. Start from one process that hurts, connect it to the real systems, agree the number and put a date on it. That’s the whole method. It isn’t magic; it’s engineering.",[421,422,423],"prose-footnotes",{},[15,424,425],{},"¹ MIT NANDA, “The GenAI Divide: State of AI in Business 2025”.",{"title":203,"searchDepth":204,"depth":204,"links":427},[428,429,430,431,432],{"id":259,"depth":204,"text":260},{"id":274,"depth":204,"text":275},{"id":301,"depth":204,"text":302},{"id":389,"depth":204,"text":390},{"id":415,"depth":204,"text":416},"strategy","2026-09-22","In 2025, MIT’s NANDA initiative looked at how companies were using generative AI and found that about 95% of enterprise pilots delivered no measurable impact on the P&L.1 Billions spent, demos applauded, and almost nothing changed in how the business actually ran.","Why 95% of AI pilots never reach ==production.==","95",{},"\u002Finsights\u002Fwhy-95-percent-of-ai-pilots-never-reach-production",7,[442,443,444],"what-an-agentic-os-is-and-what-it-isnt","every-agent-needs-a-judge","the-30-day-playbook-week-by-week",{"title":241,"description":435},"insights\u002Fwhy-95-percent-of-ai-pilots-never-reach-production","It’s rarely the model. It’s the data, the owner and the missing number. Here’s what the ones that ship do differently.",[449,450,451,452,453],{"id":259,"label":260},{"id":274,"label":275},{"id":301,"label":302},{"id":389,"label":390},{"id":415,"label":416},"9wJDwNO8uXgEwGK-cE1WYthNEN-J5Ic1e7kI_crM30g",{"id":456,"title":457,"author":458,"blobHue":10,"body":459,"category":212,"date":678,"description":463,"draft":214,"extension":215,"featured":214,"headline":679,"hue":680,"letter":681,"meta":682,"navigation":220,"path":683,"readMinutes":222,"related":684,"seo":685,"stem":686,"summary":687,"toc":688,"__hash__":694},"insights_en\u002Finsights\u002Fwhat-agents-should-never-do-alone.md","What agents should never do alone",{"name":7,"role":8,"bio":9},{"type":12,"value":460,"toc":671},[461,464,467,471,477,483,489,492,496,499,502,505,509,512,532,535,539,542,650,653,658,662,665,668],[15,462,463],{},"The question we hear most from operations leads is not “can the agent do this?”. It is “what stops it doing something stupid?”. The honest answer is not a better prompt. It is a short list of rules, written by the people who own the risk, that the agent cannot talk its way around.",[15,465,466],{},"Writing those rules is less work than it sounds. Almost everything that should stay with a person falls into three kinds of decision.",[22,468,470],{"id":469},"three-kinds-of-decisions","Three kinds of decisions",[15,472,473,476],{},[46,474,475],{},"Money."," Refunds, credits, discounts, prices, payments. Anything that moves money in or out of the business, or changes what a customer pays.",[15,478,479,482],{},[46,480,481],{},"Promises."," A delivery date, a replacement, an exception to the policy, anything that commits the company to a customer or a supplier. A wrong answer is fixable. A wrong promise has to be honoured or broken.",[15,484,485,488],{},[46,486,487],{},"Anything you can’t undo."," Deleting records, sending to your whole customer base, cancelling or changing an order that is already moving. If the mistake cannot be reversed by clicking something, a person decides.",[15,490,491],{},"Everything else, like reading, classifying, drafting, looking up, summarising and routing, is usually safe for an agent to do alone, provided it is visible afterwards.",[22,493,495],{"id":494},"draft-then-apply","Draft, then apply",[15,497,498],{},"The most useful distinction in approval rules is between preparing an action and executing it. An agent can do all the work of a refund: find the order, check the policy, calculate the amount, write the reply. Then it stops, and a person applies it with one tap.",[15,500,501],{},"This keeps most of the time saving and almost none of the risk. The person is not doing the work. They are checking a finished proposal, with the order, the policy and the amount on one screen.",[15,503,504],{},"Over time, some drafts earn the right to apply themselves. When a person has approved the same kind of action enough times without changes, you can move it below the threshold. That is a decision for the team that owns it, not for the agent.",[22,506,508],{"id":507},"thresholds-not-feelings","Thresholds, not feelings",[15,510,511],{},"“Ask me when it’s important” is not a rule. An agent cannot know what feels important to you. A rule needs a number, a category or a list:",[396,513,514,520,526],{},[43,515,516,519],{},[46,517,518],{},"A number."," Refunds up to a limit are applied; above it, they go to the ops lead.",[43,521,522,525],{},[46,523,524],{},"A category."," Any change to a contract goes to legal, whatever the value.",[43,527,528,531],{},[46,529,530],{},"A list."," New suppliers, key accounts and anything to the press always go to a person.",[15,533,534],{},"Every rule names who decides and where they are asked: in Slack, Teams or email, wherever that person already works. A rule that sends approvals to a dashboard nobody opens is a rule that stops the business.",[22,536,538],{"id":537},"a-worked-example","A worked example",[15,540,541],{},"This is the shape of a rule set for an online store’s support and commerce agents. The limits are yours to set; the structure is what matters.",[95,543,544,557],{},[98,545,546],{},[101,547,548,551,554],{},[104,549,550],{},"ACTION",[104,552,553],{},"AGENT ALONE",[104,555,556],{},"GOES TO A PERSON",[114,558,559,572,585,598,611,624,637],{},[101,560,561,566,569],{},[119,562,563],{},[46,564,565],{},"Order status, tracking, policy questions",[119,567,568],{},"Answers",[119,570,571],{},"Never, unless the customer asks for one",[101,573,574,579,582],{},[119,575,576],{},[46,577,578],{},"Refund",[119,580,581],{},"Drafts every one, applies small ones",[119,583,584],{},"Above €500 in this example, the ops lead",[101,586,587,592,595],{},[119,588,589],{},[46,590,591],{},"Voucher or discount",[119,593,594],{},"Proposes, inside the margin floor",[119,596,597],{},"Anything outside the agreed limits",[101,599,600,605,608],{},[119,601,602],{},[46,603,604],{},"Price change",[119,606,607],{},"Never writes a price",[119,609,610],{},"The pricing owner, after the margin check",[101,612,613,618,621],{},[119,614,615],{},[46,616,617],{},"Delivery date",[119,619,620],{},"Quotes what the courier data says",[119,622,623],{},"Any exception or guarantee",[101,625,626,631,634],{},[119,627,628],{},[46,629,630],{},"Order change or cancellation",[119,632,633],{},"Drafts",[119,635,636],{},"Always applied by a person",[101,638,639,644,647],{},[119,640,641],{},[46,642,643],{},"Campaign to the whole base",[119,645,646],{},"Prepares and simulates",[119,648,649],{},"Always sent by a person",[15,651,652],{},"Start with the right-hand column. The right-hand column is the list of things your team has decided to keep. Everything to its left is work they no longer have to do.",[267,654,655],{},[15,656,657],{},"If you can’t undo it, a person does it.",[22,659,661],{"id":660},"the-veto-that-isnt-a-model","The veto that isn’t a model",[15,663,664],{},"Some rules are too important to leave to a language model, even one that is being checked. Margin is the clearest case.",[15,666,667],{},"In the marketplace system we built, no agent ever writes a price. Agents can propose a discount or a clearance offer, but every proposal passes through a margin guardrail first. The guardrail is plain code, not a model. It calculates the floor for that product after returns, fees and shipping, and anything below it is vetoed. There is nothing to persuade and no prompt to get wrong.",[15,669,670],{},"That is the pattern we use wherever a rule can be expressed as arithmetic or a lookup. The model proposes, deterministic code checks, and a person approves what the rules say a person approves. Each layer does the thing it is good at, and none of them is asked to be the last line of defence alone.",{"title":203,"searchDepth":204,"depth":204,"links":672},[673,674,675,676,677],{"id":469,"depth":204,"text":470},{"id":494,"depth":204,"text":495},{"id":507,"depth":204,"text":508},{"id":537,"depth":204,"text":538},{"id":660,"depth":204,"text":661},"2026-09-08","What agents should never do ==alone.==","sun","!",{},"\u002Finsights\u002Fwhat-agents-should-never-do-alone",[443,444,442],{"title":457,"description":463},"insights\u002Fwhat-agents-should-never-do-alone","A practical way to write approval rules: money, promises and anything you can’t undo.",[689,690,691,692,693],{"id":469,"label":470},{"id":494,"label":495},{"id":507,"label":508},{"id":537,"label":538},{"id":660,"label":661},"mWnCbu3ZKOI9CJrpjxbakDDnQwkJuenNM8EI7HSjdMA",{"id":696,"title":697,"author":698,"blobHue":10,"body":699,"category":433,"date":924,"description":703,"draft":214,"extension":215,"featured":214,"headline":925,"hue":926,"letter":927,"meta":928,"navigation":220,"path":929,"readMinutes":222,"related":930,"seo":931,"stem":932,"summary":933,"toc":934,"__hash__":940},"insights_en\u002Finsights\u002Fbuild-buy-or-both.md","Build, buy or both?",{"name":7,"role":8,"bio":9},{"type":12,"value":700,"toc":917},[701,704,707,711,714,721,728,734,737,741,744,770,773,777,780,786,792,872,875,880,884,887,890,894,914],[15,702,703],{},"Every company we talk to already pays for some AI. A writing assistant here, a meeting summariser there, a chatbot on the website. The question is rarely whether to use AI. It is which work deserves more than a subscription.",[15,705,706],{},"The honest answer is that most of it doesn’t. And the part that does usually needs both: things you buy and things you build.",[22,708,710],{"id":709},"when-buying-is-enough","When buying is enough",[15,712,713],{},"An off-the-shelf tool is the right choice when three things are true.",[15,715,716,717,720],{},"The work is ",[46,718,719],{},"generic",". Summarising a meeting, rewriting an email, translating a document, drafting a first version of a job ad. Every company does it roughly the same way, so a product built for everyone fits you well.",[15,722,723,724,727],{},"It needs ",[46,725,726],{},"no live data",". The input is whatever the person pastes in, and the output goes back to that person. Nothing has to be read from your ERP or written into your CRM.",[15,729,730,733],{},[46,731,732],{},"Nobody owns an outcome."," It makes individuals faster, but no team has signed up to move a number with it. If it gets worse next month, someone switches tools and nothing breaks.",[15,735,736],{},"For this kind of work, building your own is a waste of money. Buy the tool, set some sensible rules on what data goes into it, and move on.",[22,738,740],{"id":739},"when-a-process-deserves-its-own-agent","When a process deserves its own agent",[15,742,743],{},"The picture changes when the work runs through your systems and your rules. A process deserves its own agent when:",[40,745,746,752,758,764],{},[43,747,748,751],{},[46,749,750],{},"It lives in your systems."," The agent has to read the order, the ticket history, the supplier file or the price list, and write back to them.",[43,753,754,757],{},[46,755,756],{},"There is a number."," First response time, hours spent on admin, stock left unsold. A team owns it and wants it to move.",[43,759,760,763],{},[46,761,762],{},"Some decisions need a person."," Refunds, promises, anything that cannot be undone. Those need approval rules that match how your company works, not how a vendor imagined companies work.",[43,765,766,769],{},[46,767,768],{},"It is specific to you."," Your policies, your exceptions, your tone, the thing only Maria knows. A generic tool does not know any of it and you cannot teach it properly.",[15,771,772],{},"No product sold to everyone can do all four for you. It doesn’t know your systems, it can’t own your number, and its rules are someone else’s.",[22,774,776],{"id":775},"the-both-pattern","The both pattern",[15,778,779],{},"In practice, the answer for most serious processes is both.",[15,781,782,785],{},[46,783,784],{},"You buy the commodity."," The language models themselves, from the big model providers. Hosting, email delivery, search, the plumbing. Nobody should build their own model to answer supplier emails.",[15,787,788,791],{},[46,789,790],{},"You build what is yours."," The agents, the approval rules and the shared memory. That is where your process, your policies and your judgement live, and it is the part that makes the difference between a demo and a system you can rely on.",[95,793,794,806],{},[98,795,796],{},[101,797,798,800,803],{},[104,799],{},[104,801,802],{},"BUY",[104,804,805],{},"BUILD",[114,807,808,821,833,846,859],{},[101,809,810,815,818],{},[119,811,812],{},[46,813,814],{},"Models",[119,816,817],{},"From the model providers",[119,819,820],{},"Never",[101,822,823,828,831],{},[119,824,825],{},[46,826,827],{},"Infrastructure",[119,829,830],{},"Hosting, email, storage",[119,832,820],{},[101,834,835,840,843],{},[119,836,837],{},[46,838,839],{},"Agents",[119,841,842],{},"Rarely a fit",[119,844,845],{},"Built around your process",[101,847,848,853,856],{},[119,849,850],{},[46,851,852],{},"Approval rules",[119,854,855],{},"Generic settings",[119,857,858],{},"Written with your team",[101,860,861,866,869],{},[119,862,863],{},[46,864,865],{},"Shared memory",[119,867,868],{},"Empty until you fill it",[119,870,871],{},"Your policies, history and decisions",[15,873,874],{},"This split also protects you. Models improve and prices change every few months. When the agents, rules and memory are yours, swapping the model underneath is an engineering task, not a migration. The marketplace system we built uses models from several providers at once, each chosen for the job, and the system does not depend on any single one.",[267,876,877],{},[15,878,879],{},"Buy the model. Build the part that knows your business.",[22,881,883],{"id":882},"who-owns-what","Who owns what",[15,885,886],{},"Whatever you build, ask who owns it at the end. In our projects the answer is simple. The agents run in your environment and the memory is in your systems. You own the code, the prompts and the data, so you can run them without us.",[15,888,889],{},"That matters more than it seems. A process that runs on an agent you do not own is a process someone else can reprice, change or switch off.",[22,891,893],{"id":892},"a-quick-test","A quick test",[392,895,897],{"title":896},"Before you buy another AI tool, ask",[396,898,899,902,905,908,911],{},[43,900,901],{},"Does it need to read or write your live systems?",[43,903,904],{},"Is there a number that a team is expected to move with it?",[43,906,907],{},"Are there decisions in it that must stay with a person?",[43,909,910],{},"Does it depend on policies, exceptions or tone that are specific to you?",[43,912,913],{},"If the vendor changed its price or product tomorrow, would a process stop?",[15,915,916],{},"If every answer is no, buy it. If two or more are yes, the process probably deserves its own agent, built on bought models, with rules and memory that belong to you.",{"title":203,"searchDepth":204,"depth":204,"links":918},[919,920,921,922,923],{"id":709,"depth":204,"text":710},{"id":739,"depth":204,"text":740},{"id":775,"depth":204,"text":776},{"id":882,"depth":204,"text":883},{"id":892,"depth":204,"text":893},"2026-08-04","Build, buy or ==both?==","violet","?",{},"\u002Finsights\u002Fbuild-buy-or-both",[442,224,444],{"title":697,"description":703},"insights\u002Fbuild-buy-or-both","When an off-the-shelf AI tool is enough, and when your process deserves its own agent.",[935,936,937,938,939],{"id":709,"label":710},{"id":739,"label":740},{"id":775,"label":776},{"id":882,"label":883},{"id":892,"label":893},"XoWrr_D4QkTedkkrnrgccPyPPGBK8-ZFnFJfkxogvLc",1790618465045]