[{"data":1,"prerenderedAt":593},["ShallowReactive",2],{"insight-\u002Finsights\u002Fprocess\u002Fthe-rise-of-the-generalists":3,"insight-surround-\u002Finsights\u002Fprocess\u002Fthe-rise-of-the-generalists":211},{"id":4,"title":5,"author":6,"authorRole":7,"body":8,"date":197,"description":14,"extension":198,"featured":199,"heroImage":197,"image":183,"imageAlt":197,"kind":200,"meta":201,"navigation":202,"path":203,"practice":204,"readingTime":205,"seo":206,"stem":208,"tags":209,"__hash__":210},"insights\u002Finsights\u002Fprocess\u002Fthe-rise-of-the-generalists.md","The Rise of the Generalists","Tom Santos","President and CEO",{"type":9,"value":10,"toc":182},"minimark",[11,15,18,21,26,29,32,35,38,42,45,52,58,61,64,68,71,76,79,83,86,90,93,96,100,103,106,109,113,116,119,122,129,133,136,139,142,145],[12,13,14],"p",{},"How many people does it take to build enterprise software?",[12,16,17],{},"For twenty years, the answer was seven or more: a product manager, a business analyst, a business stakeholder, an architect, a designer, a QA tester, and a support engineer — plus the developers actually writing code. Shipping anything meaningful took a quarter, often two.",[12,19,20],{},"This article explains what happened to the other six chairs, because the answer matters to anyone who builds software, buys software, or is planning a career in either.",[22,23,25],"h2",{"id":24},"the-old-assembly-line","The Old Assembly Line",[12,27,28],{},"The traditional model was an assembly line of specialists, and every station produced something for the next station to consume.",[12,30,31],{},"The business stakeholder brought the need and the budget. The product manager turned that into a roadmap and requirements. The business analyst turned requirements into detailed specs and acceptance criteria. The architect turned specs into technical design. The designer turned them into screens and flows. Developers turned all of it into code. QA turned code into bug reports, which went back to developers. And once it shipped, a support engineer triaged everything the process missed — forever.",[12,33,34],{},"Count the arrows in that chain. Every arrow is a handoff, and every handoff is a translation. Meaning leaks at each one. I spent over a decade working with companies and watching this play out: a requirement leaves a stakeholder's mouth, passes through three translations, and arrives in the codebase as something the stakeholder never asked for. Then the whole chain runs in reverse to fix it. The person who understood the problem was always four steps removed from the person typing the solution.",[12,36,37],{},"That structure existed for one reason: no single human could hold all seven jobs at once. So we parallelized, and we paid for parallelization with coordination — meetings, tickets, specs, status calls. On most enterprise projects, coordination consumed more calendar time than construction did.",[22,39,41],{"id":40},"the-new-org-chart-the-builder-and-the-architect","The New Org Chart: The Builder and The Architect",[12,43,44],{},"Agentic development collapses that chain into two roles.",[12,46,47,51],{},[48,49,50],"strong",{},"The Builder"," owns the problem end to end. The Builder talks directly to the stakeholder, shapes the requirement, directs the agents that write the code, and verifies the output against the original intent. The Builder is the product manager, the business analyst, and the developer collapsed into one person — because when agents handle construction, there is nothing left to parallelize, and the translation layers between those roles lose their purpose. The person who understands the problem is now the person shipping the solution.",[12,53,54,57],{},[48,55,56],{},"The Sr. Architect"," validates the architecture the agents propose. Agents are extremely competent architects. In the new model, the agents still design the system and its features — they propose the data model, the service boundaries, the integration approach. What they need is a professional validator. A Builder without an architecture background has no way to judge whether a proposed approach is the right one for this system, this tech stack, and the next five years of its life. The Sr. Architect does.",[12,59,60],{},"The stakes here are foundational. A system's architecture is its backbone. Get it wrong early, and every enhancement and every support ticket for years afterward becomes a nightmare. The Sr. Architect is the technical generalist standing guard over that backbone: someone who understands coding principles, architecture standards across varied tech stacks, and CI\u002FCD. In practice, these are engineers who moved into the role as demand for human developers declined. Their implementation years became their qualification for judging implementations they never have to write.",[12,62,63],{},"That is the nucleus of the new application development team. Two humans, and a harness that carries the institutional knowledge of the roles that used to sit between them.",[22,65,67],{"id":66},"where-the-other-five-roles-went","Where the Other Five Roles Went",[12,69,70],{},"This is the question that decides whether you trust this model, so I will answer it directly. The work went to three places.",[72,73,75],"h3",{"id":74},"absorbed-into-the-builder","Absorbed into the Builder",[12,77,78],{},"Product management, business analysis, and stakeholder liaison work still happen — the Builder does them directly. These roles existed to move understanding between people. When one person carries the understanding from conversation to production, the roles merge into that person.",[72,80,82],{"id":81},"encoded-into-the-harness","Encoded into the Harness",[12,84,85],{},"Architecture review, QA, and design standards moved into automated contracts and gates. And here is the number that surprises people: in our builds, verification effort now runs at roughly four times construction effort — a Build:Verify ratio approaching 1:4. Quality assurance grew. It moved from a person running test plans at the end of the line to a system enforcing standards at every step of it.",[72,87,89],{"id":88},"automated-by-agents","Automated by Agents",[12,91,92],{},"Test generation, regression runs, documentation, and first-line support triage are now agent work, executed continuously instead of scheduled into sprints.",[12,94,95],{},"The work survived. The coordination died. Seven people's worth of output still gets produced — by two people directing systems that carry the other five roles' knowledge.",[22,97,99],{"id":98},"what-this-means-for-speed","What This Means for Speed",[12,101,102],{},"The speed gain comes from removing handoffs, not from typing faster.",[12,104,105],{},"Agents write code quickly, but that is the smaller effect. The larger effect is that nobody waits for anybody. There is no ticket queue between the requirement and the build, no sprint boundary between the build and the test, no escalation path between the bug and the fix. It took 6 months and 1 Builder to build Luna, our cutting edge Agent Workforce platform. The old model would have quoted a team of eight and 2+ years for the same scope, and most of that time would have been spent in the arrows between the boxes, not in the boxes themselves.",[12,107,108],{},"Days replace quarters because the human team got smaller, and the human team got smaller because translation became unnecessary.",[22,110,112],{"id":111},"what-this-means-for-you","What This Means for You",[12,114,115],{},"If you are wondering where you fit in this model, the skills that matter have shifted.",[12,117,118],{},"A good Builder needs product judgment and domain fluency more than syntax. If you can effectively state a problem, define what \"done\" looks like, anticipate the edge cases, then you can direct a build. The scarce skill is knowing what to build and recognizing when the output is wrong.",[12,120,121],{},"A good Sr. Architect needs breadth over depth: coding principles, architecture standards across stacks, and CI\u002FCD fluency, combined with the judgment to look at an agent's proposed design and say \"yes, this holds up for five years\" or \"no, this collapses at scale.\" The leverage multiplied — one validator's judgment now shapes every system that ships, and the engineering career path leads here as human implementation demand declines.",[12,123,124,125,128],{},"And an honest word about the transition: the seven-role model employed a lot of people, and this model employs fewer per product. The people who thrive in it are the ones from those collapsed roles who move toward ",[48,126,127],{},"ownership"," — the BA who becomes a Builder, the QA lead who becomes the Architect of the verification gates. The specialist chairs are disappearing. Welcome to the rise of the Generalists.",[22,130,132],{"id":131},"the-answer","The Answer",[12,134,135],{},"So, how many people does it take to build enterprise software?",[12,137,138],{},"Two. A Builder who owns the problem and a Sr. Architect who guards the backbone. And you scale that base team up based on complexity and workload. ",[12,140,141],{},"We are proving this daily at Lunari. If you want to see what this new way of working ships, follow along.",[143,144],"hr",{},[12,146,147,151],{},[148,149,150],"em",{},"Lunari builds agentic enterprise platforms and advises leadership teams deploying them. If this maps to a problem you are working on, ",[148,152,153,181],{},[48,154,155],{},[156,157,180],"a",{"className":158,"href":177,"rel":178},[159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176],"text-primary","border-b","border-transparent","hover:border-primary","font-medium","rounded-xs","outline-primary\u002F25","focus-visible:outline-3","transition-colors","focus-visible:has-[>code]:outline-0","[&>code]:border-dashed","[&>code]:outline-primary\u002F25","focus-visible:[&>code]:outline-3","hover:[&>code]:border-primary","hover:[&>code]:text-primary","focus-visible:[&>code]:border-primary","focus-visible:[&>code]:text-primary","[&>code]:transition-colors","https:\u002F\u002Fwww.lunari-inc.com\u002Fcontact",[179],"nofollow","we should talk","****.",{"title":183,"searchDepth":184,"depth":184,"links":185},"",3,[186,188,189,194,195,196],{"id":24,"depth":187,"text":25},2,{"id":40,"depth":187,"text":41},{"id":66,"depth":187,"text":67,"children":190},[191,192,193],{"id":74,"depth":184,"text":75},{"id":81,"depth":184,"text":82},{"id":88,"depth":184,"text":89},{"id":98,"depth":187,"text":99},{"id":111,"depth":187,"text":112},{"id":131,"depth":187,"text":132},null,"md",false,"article",{},true,"\u002Finsights\u002Fprocess\u002Fthe-rise-of-the-generalists","advisory",10,{"title":207,"description":14},"How Many Humans Does It Take to Ship an Application in 2026? file","insights\u002Fprocess\u002Fthe-rise-of-the-generalists",[],"aS_CyFWhqf1fS0jkkGdlYwOUW0ui2mSXCU_klnwvRlQ",[212,366,483],{"id":213,"title":214,"author":215,"authorRole":216,"body":217,"date":353,"description":354,"extension":198,"featured":199,"heroImage":197,"image":197,"imageAlt":197,"kind":200,"meta":355,"navigation":202,"path":356,"practice":204,"readingTime":357,"seo":358,"stem":359,"tags":360,"__hash__":365},"insights\u002Finsights\u002Fmaking-agents-accountable.md","Accountability Is the Hard Part of Agentic Engineering","Lunari","Founding Team",{"type":9,"value":218,"toc":342},[219,222,226,229,232,238,242,246,249,256,260,263,266,270,277,280,284,287,291,294,301,305,308,324,327,329],[12,220,221],{},"Every enterprise AI pilot demos well. A striking number never reach production. The\ngap between the two is the absence of the machinery that makes an autonomous\nactor safe to trust with consequential work.",[22,223,225],{"id":224},"the-demo-to-production-gap","The Demo-to-Production Gap",[12,227,228],{},"A demo is a single happy path executed once, under supervision, on curated data. Production\nis thousands of executions across months, on data nobody cleaned, with staff turnover,\nchanging permissions, and an auditor who will eventually ask what happened on a specific\nTuesday in March.",[12,230,231],{},"Those are different engineering problems. The first needs a capable model. The second needs\ninfrastructure that most organizations have not built, because until recently there was no\nreason to.",[233,234,235],"blockquote",{},[12,236,237],{},"An agent that cannot be audited cannot be trusted with work that matters. Accountability\nis the substrate everything else stands on, and it has to be there from the start.",[22,239,241],{"id":240},"four-things-that-have-to-exist","Four Things That Have to Exist",[72,243,245],{"id":244},"memory-that-persists","Memory That Persists",[12,247,248],{},"Context windows are working memory, not institutional memory. When a session ends, an agent\nwithout persistent storage forgets everything — the decision it made, the reason it made it,\nand the correction you gave it. Every session restarts from zero.",[12,250,251,252,255],{},"Real memory means facts that survive across sessions, models, and vendors, with the\nprovenance of each belief attached. It also means being able to reconstruct what the agent\nbelieved at a past moment, separately from what it believes now — because when something\ngoes wrong, the question is always ",[148,253,254],{},"what did it know at the time",".",[72,257,259],{"id":258},"identity-that-travels","Identity That Travels",[12,261,262],{},"If an agent's identity lives inside one vendor's product, you do not have an agent — you\nhave a subscription. The role, its accumulated knowledge, and its track record should be\nportable across the models that execute it.",[12,264,265],{},"This matters commercially as much as technically. Model capability is converging and prices\nare falling. Betting your operating model on a specific provider's proprietary agent\nabstraction is a strategic risk as much as an architectural one.",[72,267,269],{"id":268},"permission-that-actually-holds","Permission That Actually Holds",[12,271,272,273,276],{},"Most agentic frameworks treat permissions as a prompt instruction: ",[148,274,275],{},"do not modify records\nwithout approval",". That is a request, made to a non-deterministic system, that will\neventually be declined.",[12,278,279],{},"Real permissioning is enforced outside the model, at the point of execution. The agent\nproposes; a deterministic layer decides whether the action is allowed, given who the agent\nis acting for and what it has been authorized to do. The model cannot talk its way past a\nsystem it does not control.",[72,281,283],{"id":282},"audit-that-stands-up","Audit That Stands Up",[12,285,286],{},"When an agent takes an action, six months later someone will need to know: what did it do,\non whose authority, based on what belief, and could it have been prevented. If the answer\nlives in an ephemeral chat log, the organization cannot defend the decision.",[22,288,290],{"id":289},"earned-autonomy","Earned Autonomy",[12,292,293],{},"The pattern we keep returning to is incremental trust. An agent begins with narrow scope and\nsupervised execution. As it demonstrates reliability on a class of work, its scope widens.\nFailures narrow it again.",[12,295,296,297,300],{},"This mirrors how organizations onboard people, and for the same reason: capability has to\nbe demonstrated before it is relied on. It also gives leadership a defensible answer to the question\nthat stalls most deployments — ",[148,298,299],{},"how do you know it will not do something catastrophic?"," The\nanswer is that it has not been granted the ability to, and will not be until it has earned it.",[22,302,304],{"id":303},"what-this-means-practically","What This Means Practically",[12,306,307],{},"If you are evaluating agentic systems, the questions worth asking are less about the model\nand more about the substrate:",[309,310,311,315,318,321],"ul",{},[312,313,314],"li",{},"Where does the agent's memory live, and who can audit it?",[312,316,317],{},"If you changed model providers next quarter, what would you lose?",[312,319,320],{},"Are permissions enforced in code, or requested in a prompt?",[312,322,323],{},"Can you reconstruct why the agent acted as it did, months after the fact?",[12,325,326],{},"A vendor with strong answers to those has built infrastructure. A vendor lacking them has\nbuilt a demo, and closing that gap will consume the budget you allocated for results.",[143,328],{},[12,330,331,334],{},[148,332,333],{},"Lunari builds agentic enterprise platforms and advises leadership teams deploying them.\nIf this maps to a problem you are working on, ",[148,335,336,341],{},[48,337,338],{},[156,339,180],{"href":340},"\u002Fcontact","**.",{"title":183,"searchDepth":184,"depth":184,"links":343},[344,345,351,352],{"id":224,"depth":187,"text":225},{"id":240,"depth":187,"text":241,"children":346},[347,348,349,350],{"id":244,"depth":184,"text":245},{"id":258,"depth":184,"text":259},{"id":268,"depth":184,"text":269},{"id":282,"depth":184,"text":283},{"id":289,"depth":187,"text":290},{"id":303,"depth":187,"text":304},"August 9, 2026","Model capability stopped being the bottleneck. What blocks agents from real enterprise work is everything around the model — memory, identity, permission, and audit. Here is how we think about building it.",{},"\u002Finsights\u002Fmaking-agents-accountable",8,{"title":214,"description":354},"insights\u002Fmaking-agents-accountable",[361,362,363,364],"agentic engineering","governance","enterprise AI","subcortex","fazrZ7kdzt8tyICgZpa09iiqIgwaZr9LzRvtpmAs6dk",{"id":367,"title":368,"author":6,"authorRole":7,"body":369,"date":471,"description":472,"extension":198,"featured":202,"heroImage":197,"image":473,"imageAlt":474,"kind":200,"meta":475,"navigation":202,"path":476,"practice":204,"readingTime":477,"seo":478,"stem":479,"tags":480,"__hash__":482},"insights\u002Finsights\u002Fprocess\u002Fthe-question-that-decides-whether-your-agentic-strategy-fails.md","The Question That Decides Whether Your Agentic Strategy Fails",{"type":9,"value":370,"toc":467},[371,374,377,380,384,387,431,434,437,440,443,447,450,453,456,459,461],[12,372,373],{},"If you're getting ready to build with agents or if you're struggling to see the results you keep hearing about, let me ask you the one question that will determine whether your agentic engineering strategy works:",[12,375,376],{},"Are you starting with the process, or with the model?",[12,378,379],{},"I've watched both paths play out. I spent more than a decade consulting on exactly this, inside startups, mid-market companies, and global enterprises, brought in to fix processes that were producing bad outcomes. The pattern held across every engagement: the fix lived in one of a handful of core phases, and it had nothing to do with the platform they'd paid for or the framework they'd just rolled out. Today my teams ship applications and features through an AI-PDLC every day, and I can tell you the choice you make in your first week will shape everything that follows.",[22,381,383],{"id":382},"the-process-pieces-have-never-changed","The Process Pieces Have Never Changed",[12,385,386],{},"Here's what's true about application development: Frameworks have cycled for decades. Waterfall gave way to Agile, Agile spawned a dozen flavors, while the durable phases underneath stayed put:",[388,389,390,396,402,408,414,420,425],"ol",{},[312,391,392,395],{},[48,393,394],{},"ENVISION"," the need.",[312,397,398,401],{},[48,399,400],{},"GATHER"," the requirements that solve the correct problem.",[312,403,404,407],{},[48,405,406],{},"DESIGN"," the concept and design the architecture.",[312,409,410,413],{},[48,411,412],{},"BUILD"," the product.",[312,415,416,419],{},[48,417,418],{},"TEST"," and verify the product.",[312,421,422,413],{},[48,423,424],{},"RELEASE",[312,426,427,430],{},[48,428,429],{},"SUPPORT"," the users.",[12,432,433],{},"Frameworks decide how you slice those phases and how fast you loop them. The phases themselves are the work, and they form a chain. Your outcome is governed by the weakest link in it.",[12,435,436],{},"Some teams did improve when they moved to Agile, because long feedback loops were the wrong fit for products with uncertain requirements. The framework change fixed a fit problem, and that's real. What it left untouched was practice. Teams that ran thin requirements and rushed design under Waterfall carried those habits into their sprints, and the work broke down for the same reasons it broke down before.",[12,438,439],{},"Many of those teams are looking at agentic development now and wondering why they're still struggling, and the truth is agents amplify whatever process you hand them — good and bad. Give them a strong specification and they produce work you can ship. Give them a vague ticket and they produce volume you have to untangle.",[12,441,442],{},"And agents change the economics of the chain. Under human teams, build carried most of the cost, which is why requirements and testing were the first things cut under a deadline. When build collapses toward zero, the cost moves to specification and verification, precisely the links weak teams were already skipping.",[22,444,446],{"id":445},"start-with-the-process","Start With the Process",[12,448,449],{},"So the work starts before an agent is ever engaged. Take the chain and audit it honestly, phase by phase. Strengthen the chain, and you strengthen your agentic output.",[12,451,452],{},"Tokens accelerate a process, but they also inherit its gaps. A team that struggles to elicit real user needs will generate elegant solutions to imagined problems. A team that skips technical design will accumulate architecture nobody can defend and potentially a nightmare to support. A team without a verification strategy will ship faster and learn about the consequences from its users.",[12,454,455],{},"The tools and frameworks will keep changing, but the key phases of any application development process haven't changed in decades. Your results will keep tracking the weakest links in your chain. So find them.",[12,457,458],{},"Start with the process.",[143,460],{},[12,462,463],{},[148,464,150,465,255],{},[156,466,180],{"href":340},{"title":183,"searchDepth":184,"depth":184,"links":468},[469,470],{"id":382,"depth":187,"text":383},{"id":445,"depth":187,"text":446},"2026-08-17","Agents amplify whatever process you hand them. After a decade spent fixing broken delivery processes, one question predicts whether an agentic strategy will produce results — are you starting with the process, or with the model?","\u002Fimages\u002Finsights\u002Fprocess-before-model.jpg","The seven durable phases of software delivery — envision, gather requirements, design, build, test, release, support — drawn as a connected chain of gears",{},"\u002Finsights\u002Fprocess\u002Fthe-question-that-decides-whether-your-agentic-strategy-fails",5,{"title":368,"description":472},"insights\u002Fprocess\u002Fthe-question-that-decides-whether-your-agentic-strategy-fails",[481],"process","NQ1qgUmizlV_WvHA568NpuzrAM2O-co-59_jH46PY40",{"id":4,"title":5,"author":6,"authorRole":7,"body":484,"date":197,"description":14,"extension":198,"featured":199,"heroImage":197,"image":183,"imageAlt":197,"kind":200,"meta":590,"navigation":202,"path":203,"practice":204,"readingTime":205,"seo":591,"stem":208,"tags":592,"__hash__":210},{"type":9,"value":485,"toc":578},[486,488,490,492,494,496,498,500,502,504,506,510,514,516,518,520,522,524,526,528,530,532,534,536,538,540,542,544,546,548,550,552,556,558,560,562,564,566],[12,487,14],{},[12,489,17],{},[12,491,20],{},[22,493,25],{"id":24},[12,495,28],{},[12,497,31],{},[12,499,34],{},[12,501,37],{},[22,503,41],{"id":40},[12,505,44],{},[12,507,508,51],{},[48,509,50],{},[12,511,512,57],{},[48,513,56],{},[12,515,60],{},[12,517,63],{},[22,519,67],{"id":66},[12,521,70],{},[72,523,75],{"id":74},[12,525,78],{},[72,527,82],{"id":81},[12,529,85],{},[72,531,89],{"id":88},[12,533,92],{},[12,535,95],{},[22,537,99],{"id":98},[12,539,102],{},[12,541,105],{},[12,543,108],{},[22,545,112],{"id":111},[12,547,115],{},[12,549,118],{},[12,551,121],{},[12,553,124,554,128],{},[48,555,127],{},[22,557,132],{"id":131},[12,559,135],{},[12,561,138],{},[12,563,141],{},[143,565],{},[12,567,568,570],{},[148,569,150],{},[148,571,572,181],{},[48,573,574],{},[156,575,180],{"className":576,"href":177,"rel":577},[159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176],[179],{"title":183,"searchDepth":184,"depth":184,"links":579},[580,581,582,587,588,589],{"id":24,"depth":187,"text":25},{"id":40,"depth":187,"text":41},{"id":66,"depth":187,"text":67,"children":583},[584,585,586],{"id":74,"depth":184,"text":75},{"id":81,"depth":184,"text":82},{"id":88,"depth":184,"text":89},{"id":98,"depth":187,"text":99},{"id":111,"depth":187,"text":112},{"id":131,"depth":187,"text":132},{},{"title":207,"description":14},[],1787085544416]