Venture capital did not spend the last 12 hours chasing another giant foundation-model headline. Instead, the money moved into the systems that make AI and industrial technology usable in the real world: cloud access control, AI-agent governance, enterprise data semantics, voice infrastructure, pathogen diagnostics, PFAS destruction, radar satellites, and a fresh quantum-computing architecture. That mix matters. It suggests investors are still willing to fund ambitious science, but they increasingly want a line of sight to deployment, distribution, and defensibility.
The strongest signal of the day came from cybersecurity and AI control planes. Act Security emerged from stealth with $60 million disclosed across seed and Series A rounds, while Hush Security added a $30 million Series A to govern non-human identities and enterprise AI agents. That lines up with a broader 2026 pattern: Crunchbase says startups at the intersection of AI and security have already raised $855 million across more than 150 seed rounds this year, putting the category on pace for a record.
At the same time, several of today’s rounds were not purely financial bets. Mitsubishi Electric anchored Array Labs’ new strategic round in space-based radar. Daikin America and Veralto backed Claros in PFAS destruction. Cedars-Sinai Intellectual Property Co. co-led Vivid Dx’s seed financing in rapid diagnostics. Akamai joined Hush as a strategic investor. When corporate buyers show up this early, they are usually paying for more than upside; they are buying optionality on procurement, distribution, and technical alignment.
That dynamic also fits the broader capital backdrop. Global startup funding hit a record $510 billion in the first half of 2026, according to Crunchbase, even as the public market wrestles with the cost of financing the AI infrastructure buildout and credit analysts warn that enthusiasm around AI could outrun returns. Private investors today look less interested in open-ended exposure and more interested in companies that can meter risk, move faster in regulated settings, or sell into industries with budget urgency.
The Macro Environment: Capital Is Moving Up the Stack and Toward Control
The headline number for venture this year is still abundance. Crunchbase’s H1 data shows global startup investment reached $510 billion in the first half of 2026, already above the full-year total for 2025. But that abundance is not evenly distributed. The market has become more concentrated around AI, deep infrastructure, and startups that sit close to either the cost structure or the control structure of modern enterprises. Today’s venture funding slate is a good micro-sample of that concentration.
That concentration is being shaped by what is happening in adjacent public and project-finance markets. Meta and BlackRock announced a $14 billion joint venture for a large Texas data-center campus, while the Wall Street Journal reported that financing the AI data-center boom is getting more expensive, with record bond issuance contributing to higher funding costs. Fitch, meanwhile, warned that AI exuberance itself is becoming a meaningful global credit risk. When public-market capital gets pricier and more scrutinized, private investors tend to prefer startups that can either reduce AI risk, govern its deployment, or sell into the bottlenecks that hyperscalers and enterprises cannot avoid.
That is why control-plane companies stood out today. Act Security is attacking permission sprawl in cloud environments. Hush Security is trying to become the governance layer for the coming flood of AI agents. Credible Data is building semantic context for enterprise AI so agents do not hallucinate their way through structured data. These are not flashy consumer stories. They are the sort of infrastructure bets investors make when they believe AI adoption is real, but they also believe operational failure is where budgets will open first.
The physical side of the market is just as revealing. Claros is commercializing PFAS destruction for regulated industrial customers. Array Labs is pairing venture money with a defense-and-space incumbent to bring radar tracking services to Asia-Pacific security buyers. ZuriQ is an architectural bet that trapped-ion quantum computing still has room for a reset. CORE Biomedicine and Vivid Dx show that biotech investors remain willing to underwrite science-heavy platforms when the clinical need is plain, and the technical thesis is differentiated.
The investor psychology here is not subtle. Capital is still flowing, but it is flowing toward leverage points: governance, supply, infrastructure, regulated demand, and architectures that could produce outsized technical discontinuities. Founders should read that as a message. In this market, the best pitch is no longer “AI makes this possible.” It is “we sit where budgets become recurring and where failure is too expensive to ignore.”
Top Tech Funding Roundup
Act Security raises $60 million in funding to make cloud security about access, not alerts
Act Security is one of the clearest examples of where security money is moving. The company emerged from stealth today with $60 million in funding and a thesis that many security teams will find familiar: patching and posture management are losing ground because cloud environments have accumulated too much access sprawl, and AI can now exploit those paths faster than humans can remediate them. Act is trying to shift cloud security away from endless misconfiguration lists and toward deterministic boundaries around what humans, workloads, and AI agents can actually reach.
Investors care because this is not a marginal workflow improvement; it is an attempt to redefine the primary attack surface of the AI era. The founding team previously built Medigate, which was acquired by Claroty for $400 million, giving Act some founder-market credibility that matters in a crowded security field. Just as important, the financing stack tells a story. Team8 and Bessemer backed the seed round, while Notable Capital led the Series A. That is a classic progression for a security startup that wants to move from technical validation to enterprise-scale go-to-market.
Strategically, Act sits in a category that is likely to keep expanding as enterprise AI moves from copilots to autonomous agents with permissions in production systems. If that shift continues, security spend should tilt toward authorization, identity, and runtime enforcement instead of static scans alone. That is why today’s disclosure matters beyond the company itself. It is a bet that cloud security’s next wave will be defined by access control at machine speed.
Funding Details Startup: Act Security Investors: Team8, Bessemer Venture Partners, Hetz Ventures, Claltech, Notable Capital, Startpoint Capital, SVCI Amount Raised: $60 million disclosed today Total Raised: $60 million Funding Stage: Seed + Series A disclosed at stealth launch Funding Date: July 28, 2026 Headquarters: Tel Aviv, Israel Sector: Cybersecurity/cloud security / AI agent control
Claros Technologies raises $55 million in funding to commercialize PFAS destruction
Claros landed one of the day’s most consequential climate-and-industrial rounds, closing a $55 million Series B to push PFAS destruction into broader commercial deployment. PFAS, often called “forever chemicals,” have become one of the most politically charged and operationally expensive contamination issues facing manufacturers, utilities, remediation firms, and governments. Claros is not just offering analytics; it is pitching permanent destruction using its ClarosTechUV system while also expanding a mobile-lab business through ClarosLabs.
Why do investors care? Because this is climate tech with unusually direct budget owners. Tightening PFAS regulation has created a market where compliance, cleanup, and liability are pushing buyers toward technologies that can destroy contaminants rather than merely contain them. Claros also framed its platform as useful in sectors that still rely on PFAS for specific applications, including semiconductors, aerospace, and pharmaceuticals. That is a more pragmatic commercial strategy than simply arguing for elimination: solve the contamination problem without pretending the underlying materials disappear overnight.
The cap table is equally telling. Treehouse Family Capital led the round, but Daikin America and Veralto joined alongside foundations and asset managers. Strategic investors in industrial and environmental systems do not tend to write checks this early without seeing a route to procurement leverage or ecosystem pull. For founders in hard-tech, Claros is a useful case study: regulation can create a market, but distribution into incumbents still helps turn a regulatory tailwind into revenue.
Funding Details Startup: Claros Technologies Investors: Treehouse Family Capital, Daikin America, Veralto, Bush Foundation, Nord Asset Management, Saint Paul & Minnesota Foundation, existing investors Amount Raised: $55 million Total Raised: Not disclosed Funding Stage: Series B Funding Date: July 28, 2026 Headquarters: Minneapolis, United States Sector: Climate tech / industrial remediation / environmental infrastructure
Fish Audio raises $52 million to build voice AI for creators and enterprises
Fish Audio’s $52 million seed round is one of those financings that says as much about the stage structure of venture as it does about the company. This is technically a seed, but it looks and feels more like a mature Series A: the company says it reached $21 million in annual recurring revenue and more than 8 million users in its first year, and TechCrunch reports it has become a widely used voice platform across creators, indie developers, game studios, and enterprise customers.
The product thesis is straightforward. Fish Audio wants to own expressive, steerable voice generation across real-time text-to-speech, voice cloning, and voice agents. Investors led by Coreline Ventures and Capital Today are betting that the voice market will not be winner-take-all around a single platform. Instead, there may be room for specialist players that combine open-source adoption, developer controls, and enterprise APIs. TechCrunch notes that the company built early traction through an open-source repo and that its customer list already includes companies such as HeyGen and Sanas.
There is also a more cautious read. Voice AI is crowded, with competitors including ElevenLabs and several enterprise voice providers, and Fish Audio has already faced creator-consent questions around uploaded voices. That makes trust and rights management part of the product, not an afterthought. Still, investors backed the company aggressively because it already has usage, revenue, and a credible argument that voice is becoming application infrastructure rather than a demo feature.
Funding Details Startup: Fish Audio Investors: Coreline Ventures, Capital Today, 359 Capital, Play Time, HF0, 645 Ventures, Parable, Carya Venture Partners, Alphalist Partners, angel investors Amount Raised: $52 million Total Raised: $52 million disclosed Funding Stage: Seed Funding Date: July 28, 2026 Headquarters: Palo Alto, United States Sector: AI infrastructure/voice AI/developer platform
Hush Security raises $30 million in funding to govern the coming wave of enterprise AI agents
If Act represents the access-control side of the new security market, Hush Security represents the identity and governance side. The company raised a $30 million Series A with Akamai joining Battery Ventures and YL Ventures, bringing total funding to $41 million. Its pitch is tightly focused on the rise of non-human identities and AI agents that need permissions inside production systems but cannot be governed with the old credential model.
The timing is not accidental. Hush cites Gartner’s projection that the average Fortune 500 company could be running more than 150,000 AI agents by 2028, up from fewer than 15 last year, and Omdia’s finding that most organizations still run agents on governance models that were not built for them. The company’s answer is a central registry, no standing credentials, just-in-time permissions, continuous logging, and an emergency kill switch. That design choice puts Hush squarely in the control path for enterprise AI, which is exactly where investors want startups to sit when they fear commoditization elsewhere in the stack.
The strategic angle strengthens the case. Akamai is not a passenger investor here; it is signaling that identity and edge security will increasingly overlap with agent management. Hush also says Kyndryl has deployed the platform internally and started reselling it to enterprise customers. That combination of strategic capital and channel adjacency matters more than a vanity valuation would at this stage. It suggests the market is starting to treat AI-agent governance as a budget line, not a future problem.
Funding Details Startup: Hush Security Investors: Akamai Technologies, Battery Ventures, YL Ventures Amount Raised: $30 million Total Raised: $41 million Funding Stage: Series A Funding Date: July 28, 2026 Headquarters: Tel Aviv, Israel Sector: Cybersecurity/identity / AI agent governance
ZuriQ raises $25.5 million in funding to scale a two-dimensional trapped-ion quantum architecture
ZuriQ’s seed financing is one of the day’s purest frontier-computing bets. The ETH Zürich spinout raised $25.5 million to scale trapped-ion quantum processors built on a native two-dimensional architecture, led by Quantonation with participation from Forward.one, Extantia, Firgun Ventures, and prior investors. The company says the round builds on a $4.2 million pre-seed raised in 2025.
The technical claim is what makes the round notable. Most trapped-ion systems have been built around effectively one-dimensional chains, which can become awkward to scale. ZuriQ argues that its Penning micro-trap design, which relies on static magnetic fields, enables a natively two-dimensional structure where ions can move more freely and scale more naturally with chip area. The company says it has already demonstrated a 3×3 array of nine individually controlled ions in collaboration with ETH Zürich and with chips fabricated by Infineon.
Investors are backing an architectural reset, not incremental benchmarking. That is risky, but it is also where outsized returns in quantum will likely come from if the current front-runners hit scaling limits. In a funding market increasingly dominated by software wrappers and enterprise workflows, ZuriQ stands out as a reminder that deep-tech capital still shows up when a startup can present a plausible path from physics insight to manufacturable hardware.
Funding Details Startup: ZuriQ Investors: Quantonation, Forward.one, Extantia, Firgun Ventures, previous investors Amount Raised: $25.5 million Total Raised: About $29.7 million disclosed to date Funding Stage: Seed Funding Date: July 28, 2026 Headquarters: Zürich, Switzerland Sector: Frontier computing / quantum hardware
Array Labs raises $21 million in funding to turn space-based radar into a defense product
Array Labs’ additional $21 million round is smaller than some of the software financings today, but strategically it may carry more weight. Mitsubishi Electric anchored the oversubscribed investment and also signed a partnership with Array to bring satellite-based maritime and aircraft tracking services to defense and security customers across Asia-Pacific. That is a financing round and a route-to-market deal in the same package.
Array is building radar satellites designed to provide high-resolution, continuously updated intelligence on terrain, infrastructure, and activity, including through clouds and at night. It says it has validated formation-flying and multistatic technology and already secured R&D contracts with DARPA, the U.S. Navy, and the U.S. Air Force. In other words, this is not speculative outer-space branding; it is an attempt to build persistent sensing infrastructure with immediate defense relevance.
The round matters because it reflects a broader geopolitical shift in venture. National-security buyers increasingly want private-sector systems they can field quickly, and sovereign or quasi-sovereign industrial players want a seat at the table before those capabilities become standard. Array’s earlier $20 million Series A brought total funding to $35 million in January, implying disclosed funding now of about $56 million. That makes today’s round less about survival capital and more about strategic positioning in a defense market that is becoming regional as well as global.
Funding Details Startup: Array Labs Investors: Mitsubishi Electric, Catapult Ventures, Kompas VC, Y Combinator, prior backers Amount Raised: $21 million Total Raised: About $56 million disclosed to date Funding Stage: Strategic extension following Series A Funding Date: July 28, 2026 Headquarters: Redwood City, United States Sector: Defense tech/space infrastructure/radar intelligence
CORE Biomedicine raises $21 million in funding to advance lineage-based precision oncology
CORE Biomedicine’s $21 million Series A is one of the day’s more interesting biotech financings because it is so overtly cross-border. The company, with operations tied to Boston, Tokyo, and Suzhou, raised the round with UTokyo Innovation Platform and Elikon Venture as co-leads, joined by a roster of Japanese and Chinese investors. It also said its Japanese subsidiary received a grant from AMED, the Japan Agency for Medical Research and Development.
The company’s scientific pitch is “lineage-based” precision oncology: target core cancer drivers with more precision while widening the patient populations that can benefit. In a biotech market that often swings between platform exuberance and clinical-stage caution, this round stands out because it combines venture capital with state-linked ecosystem support. That kind of blend is increasingly common in areas where national biopharma ambitions and venture economics overlap.
For investors, the appeal is twofold. First, oncology remains one of the few biotech categories where capital can still scale around a strong mechanism thesis. Second, the company is tapping multiple innovation systems at once, which may improve access to talent, trials, and translational partnerships. For founders, CORE is a reminder that geographic complexity can be a feature if the science and market are global enough to justify it.
Funding Details Startup: CORE Biomedicine Investors: UTokyo Innovation Platform, Elikon Venture, InnoPinnacle Fund, Mitsubishi UFJ Capital, Suzhou Capital Group, CD Capital, YuanBio Venture Capital, Vision Incubate, Root Venture Partners Amount Raised: $21 million Total Raised: Not disclosed Funding Stage: Series A Funding Date: July 28, 2026 Headquarters: Boston, Tokyo, and Suzhou Sector: Biotech/precision oncology
Vivid Dx raises $15 million in funding to cut sepsis diagnosis from days to hours
Vivid Dx raised a $15 million seed round to attack one of the ugliest bottlenecks in hospital medicine: the time it takes to identify bloodstream infections and match them to effective antibiotics. The Oxford-based company says its direct-from-blood platform uses Raman spectroscopy, AI and sample preparation to identify pathogens and support phenotypic antimicrobial susceptibility testing without waiting days for culture-based workflows. Oxford Science Enterprises and Cedars-Sinai Intellectual Property Co. co-led the financing.
Investors care here because the problem is painfully concrete. In sepsis, hours matter, yet traditional diagnostics often force clinicians into broad-spectrum empiric therapy while they wait for answers. Vivid Dx says its goal is to compress that cycle from days to hours. The public-health angle is also large: its release cites research estimating that antimicrobial resistance could contribute to 39 million deaths globally between 2025 and 2050. That gives the startup a market story that combines hospital economics, patient outcomes, and antibiotic stewardship.
This round also fits a growing pattern in healthcare: investors are willing to fund AI in medicine when it is tied to a specific clinical workflow and a specific cost center. Vivid Dx is not selling generic AI to hospitals. It is selling faster decisions in one of the highest-stakes parts of acute care. That precision is usually what gets seed checks into real clinical budgets later on.
Funding Details Startup: Vivid Dx Investors: Oxford Science Enterprises, Cedars-Sinai Intellectual Property Co., Jameel Investment Management Company, additional strategic investors Amount Raised: $15 million Total Raised: Not disclosed Funding Stage: Seed Funding Date: July 28, 2026 Headquarters: Oxford, United Kingdom Sector: Biotech/diagnostics/health AI
Credible Data raises $10 million in funding to give enterprise AI a semantic backbone
Credible Data’s $10 million seed round is small by today’s standards, but strategically it may turn out to be one of the more important enterprise-AI deals in the batch. The Boulder startup is building what it calls a “trusted business context engine” for enterprise data. In plainer terms, it wants to provide the semantic layer that tells AI agents what the data in a warehouse actually means, not just where the rows and columns live. Gradient, SignalFire, and K5 Global backed the round, with angels including Godard Abel and Wes McKinney.
That thesis addresses a problem enterprise buyers have run into quickly: access to structured data is not enough if agents cannot interpret governed metrics, role-based rules, relationships, and definitions consistently. Credible Data is built on Malloy, an open-source semantic modeling language, and positions itself as the context layer that AI agents, analytics tools, and APIs can query at runtime. This is exactly the kind of infrastructure category that becomes more valuable as enterprises realize generic agent tooling breaks on company-specific data logic.
The broader significance is that venture investors are moving one layer deeper into the enterprise stack. Last year, much of the excitement was around agent interfaces. Today’s check says the market is starting to price the importance of governed semantics, auditability, and vendor-agnostic context. For founders, that is a warning and an opportunity: the next durable AI companies may be the ones that fix meaning, not just interaction.
Funding Details Startup: Credible Data Investors: Gradient, SignalFire, K5 Global, Godard Abel, Wes McKinney, Alex Dean, SV Angel, others Amount Raised: $10 million Total Raised: $10 million disclosed Funding Stage: Seed Funding Date: July 28, 2026 Headquarters: Boulder, United States Sector: Enterprise software / AI data infrastructure/developer tools
Procode raises $10 million in funding to industrialize AI medical billing for surgeons
Procode’s Series A would not usually command the same attention as a frontier-computing round, but it deserves a spot in today’s top ten because it shows where AI monetization is getting real. The company raised $10 million led by Health Velocity Capital to keep rolling up surgical billing businesses and layering on its coding engine. It says the funding follows its acquisition of The Auctus Group and a peer-reviewed study showing its fine-tuned model outperforming both human coders and general-purpose LLMs on complex surgical coding.
That matters because revenue-cycle management is one of the few areas in healthcare where ROI can be measured quickly and bought repeatedly. Procode is not selling hospitals a broad AI vision. It is selling surgeons fewer denials, better reimbursement capture, and a tighter operating system for billing teams. The startup says it now serves more than 350 plastic-surgery and dermatology providers and has raised $14 million in total venture funding.
There is also a playbook here that other vertical-AI founders should notice. Procode is combining software with acquisition strategy, using capital not only to build product but to consolidate a fragmented services market and apply AI where legacy operators already understand the workflows. That is often a more reliable route to defensibility than selling “AI automation” in the abstract.
Funding Details Startup: Procode Investors: Health Velocity Capital Amount Raised: $10 million Total Raised: $14 million Funding Stage: Series A Funding Date: July 28, 2026 Headquarters: Los Angeles, United States Sector: Healthcare software / vertical AI / revenue-cycle automation
What Today’s Funding Activity Reveals
The most obvious pattern is that AI money is moving into control layers. Act, Hush, and Credible Data are all, in different ways, trying to prevent AI systems from becoming operational liabilities. That is a notable shift from the earlier phase of the boom, when venture money rushed into models, wrappers, and interface tools. Investors now appear increasingly willing to fund companies that reduce permission risk, establish governance, or enforce meaning in enterprise data.
The second pattern is that strategic investors are showing unusually strong conviction. Mitsubishi Electric, Akamai, Daikin America, Veralto, and Cedars-Sinai Intellectual Property Co. all appeared in today’s set. That suggests many of these markets are moving from “interesting technology” to “adjacent capability we may need to buy, distribute, or integrate.” Founders often talk about strategic capital as validation. The sharper interpretation is that strategics are quietly competing not to miss the next procurement chokepoint.
The third pattern is that deep tech still attracts real seed capital when the technical wedge is unusual. Fish Audio’s $52 million seed, ZuriQ’s $25.5 million seed, Vivid Dx’s $15 million seed, and Credible Data’s $10 million seed all show that “seed” no longer means small when investors see fast traction, hard science, or a category-defining architecture. The stage labels remain early, but the check sizes increasingly reflect the cost of winning technical races before the market settles.
Geographically, the U.S. still dominates the batch, but the broader pattern is unmistakably global. Today’s list spans the United States, Israel, the United Kingdom, Switzerland, and a biopharma network stretched across Boston, Tokyo, and Suzhou. That kind of spread is common when capital is chasing either frontier research or security-adjacent applications. The money is not merely following founders; it is following ecosystems with specialized talent, policy support, or sovereign demand.
Finally, today’s rounds reinforce a simple point about demand. Buyers are not just paying for AI to sound smart. They are paying for AI to secure production systems, clean contaminated water, route radar data, reduce sepsis delays, codify enterprise data, and collect revenue more accurately. That is why this set feels healthier than a day dominated by one giant model round would have. The money is still concentrated, but at least today it flowed toward problems with visible owners and obvious budgets.
Comparative Funding Table
Startup Amount Raised Sector Funding Stage Lead Investors Country Act Security $60M Cybersecurity Seed + Series A disclosed at launch Notable Capital; Team8; Bessemer Israel Claros Technologies $55M Climate tech Series B Treehouse Family Capital United States Fish Audio $52M Voice AI Seed Coreline Ventures; Capital Today United States Hush Security $30M Cybersecurity Series A Akamai Technologies; Battery Ventures; YL Ventures Israel ZuriQ $25.5M Quantum computing Seed Quantonation Switzerland Array Labs $21M Defense tech/space Strategic extension Mitsubishi Electric United States CORE Biomedicine $21M Biotech Series A UTokyo Innovation Platform; Elikon Venture United States / Japan / China Vivid Dx $15M Diagnostics Seed Oxford Science Enterprises; Cedars-Sinai IP Co. United Kingdom Credible Data $10M Enterprise AI data infrastructure Seed Gradient; SignalFire; K5 Global United States Procode $10M Healthcare software Series A Health Velocity Capital United StatesStrategic Takeaways for Founders and Investors
For founders, the lesson from today is not that money is easy. It is that money is available when the company sits close to a control point, a compliance burden, a procurement budget, or a scientifically differentiated architecture. Act and Hush are not generic “AI security” stories; they are identity and access-control stories. Claros is not broad sustainability software; it is PFAS destruction with regulatory and industrial urgency. ZuriQ is not “quantum someday”; it is a concrete architectural argument tied to manufacturability.
For investors, strategic syndicates are becoming a stronger signal than inflated labels. Today’s best rounds paired venture firms with strategic participants who can accelerate commercial adoption: Mitsubishi Electric with Array, Akamai with Hush, Daikin and Veralto with Claros, Cedars-Sinai with Vivid Dx. In a market where public investors are increasingly questioning AI spending and credit analysts are warning about excess, distribution leverage may matter more than abstract model advantage.
Another takeaway is that enterprise buyers are beginning to pay for AI reliability, not just AI capability. Credible Data, Vivid Dx, and Procode each map to a different version of the same demand signal: enterprises and institutions want systems that produce trustworthy outputs in high-stakes settings. That is a friendlier market for startups with domain depth, workflow ownership, and measurable ROI than for startups that simply wrap general-purpose models with better UX.
The risk, however, is obvious. AI infrastructure spending remains historically intense, and the cost of financing that buildout is rising. If the public markets become less forgiving, private markets are likely to get even harsher about monetization, gross margins, and customer concentration. Founders can still raise large rounds, but today’s winners suggest they will need sharper proofs: revenue traction like Fish Audio, strategic pull like Array and Claros, or mission-critical operational value like Act, Hush, and Procode.
Conclusion
Today’s funding tape was not a referendum on hype. It was a referendum on where venture investors think friction will accumulate next. The answers were clear: AI governance, cloud access, trusted enterprise data, industrial contamination, space-based sensing, clinically actionable diagnostics, and differentiated scientific architectures. Even when consumer-facing surfaces were involved, as with Fish Audio, the real investor interest was in infrastructure, monetization, and control.
That points to a startup market that is maturing, not cooling. Capital is still abundant at the top, as the record first half of 2026 makes plain, but it is becoming more selective about where technical novelty meets budget inevitability. For founders, that means the bar is higher. For investors, it means the next durable winners may come less from grand narratives and more from companies that quietly become hard to operate without.
