The Kia Vision Meta Turismo Concept Car and the Panasonic Toughbook 56
(Credits: Kia; Panasonic)
Not to be outdone, Jason Anderson published an insightful piece on why enterprises need to be paying much more attention to the agents running in concert with their ERP systems; Mike Leone dug deep into IBM’s strategy for turning sovereignty into a product; and Melody Brue interviewed new OpenAI CMO Colin Fleming for a piece that explained what Fleming’s appointment says about OpenAI’s ambitions in the enterprise. And that’s before we get to Matt Kimball’s imminent piece on Tenstorrent’s new chip offerings for AI inference, and a significant update coming shortly from Paul Smith-Goodson about IBM’s efforts in quantum computing. (More on that topic in Paul’s “Quantum” entry in this week’s analyst roundup.) I’m proud of both the range and the depth our analysts regularly demonstrate in their work.
It’s not an accident that we also go for both range and depth in our annual Six Five Summit. We’ve recently announced that keynote speaker Marc Benioff of Salesforce will be followed by Gary Dickerson, CEO of Applied Materials. There are a lot more great speakers on the agenda, so we hope you’ll join us for this free event August 25–27 — register today!
This week, Anshel and Matt continue to take in the massive payload of news from Computex in Taipei. Mel and I are in Las Vegas for Cisco Live. Mike is attending Snowflake Summit, and Jason is at Microsoft Build — both in San Francisco. For news and hot takes from these events, be sure to follow us on X and LinkedIn.
The spring season of client and vendor events is in full swing, and we’re on the road a lot over the coming months, including at the NetApp Analyst Summit, Broadcom Mainframe Analyst Summit, AWS Analyst Summit, Databricks Data + AI Summit, HPE Discover, Pure Accelerate, and Connectivity Standards Alliance Unify — all in June. If you’ll be attending any of the same events or see that we’ll be in your city, please reach out.
Last week, Moor Insights & Strategy analysts’ perspectives appeared across major business and technology outlets including Tech Times, Bloomberg, Benzinga, Yahoo Finance, and StockTwits. Coverage highlighted AMD CEO Lisa Su’s warning that high‑bandwidth memory is emerging as the next constraint in the AI chip supply chain, key AI themes to watch at Computex 2026, and Dell’s record earnings quarter and its implications for NVIDIA’s AI leadership narrative. Analysts also weighed in on Salesforce and Snowflake earnings, where premium AI deals and large hyperscaler partnerships are reshaping expectations for cloud and enterprise software growth, and examined how a potential Maia 200 custom silicon deal between Microsoft and Anthropic could position Claude as an early proving ground for AI‑optimized chips.
Our MI&S team also published 14 deliverables — 1 Forbes Article, 1 Research Paper, 5 Research Notes, 2 Analyst Insights, and 5 Podcasts.
Check out this week’s Analyst Insights roundup for more from the MI&S team, including what’s top of mind for each analyst, our thoughts on vendor announcements, press quotes, and more.
Have a great week!
Patrick Moorhead
Models, ‘Effort,’ and the Markdown-versus-HTML Debate
The last couple of weeks at Anthropic have been interesting. The company made a huge new investment announcement and released a new model, which is being sold as something of an “incremental improvement,” plus there was some good old developer chatter.
I’ll leave the investment talk to others, but in the Opus 4.8 release I did see one very interesting item. In the release notes there was a comment about Opus 4.8’s ability to have Claude Code scale to manage hundreds of sub-agents. Personally, I believe that scalable orchestration will end up being a massive competitive threat to almost any tech services firm, and outsourcing in particular. Part of the reason companies outsource processes — or whole departments — is to hand off the complexity to another company that is supposed to be more skilled and experienced at complex project and process management. And what we are seeing so far is that AI orchestration does not scale well. It is either human-controlled and deterministic (making it brittle) or left to the model or agent, which chews up way too many tokens. So, to see Claude take that on is quite interesting, given that to scale AI we will need very smart orchestration. Oh, and one more thing: 4.87 also introduces the concept of “effort” into the mix. I want to puzzle over that some more as well.
Also in the Claude universe, there was a back-and-forth among developers about what is the better input format for Claude Code. Should we stick with .md or, as some on the Claude team suggest, move to HTML, which is much richer and easier for the human in the loop to parse and understand? Well, unsurprisingly, HTML chews up a lot more tokens than markdown, so there’s some disagreement over which is “better.” My guess is that we may see this debate end up as an optimization in future models, where (somehow or other) we will see a model trained to be far more efficient on HTML since it will provide higher-fidelity inputs, which should lead to better outcomes. We shall see.
NEW ANALYST INSIGHT: Mission-Critical ERP Needs Mission-Critical Agents
Enterprise enthusiasm for AI agents is real and accelerating — but most early deployments rest on a dangerous assumption that the IT systems beneath the agents can stay exactly as they are. In mission-critical ERP environments, where a missed compliance flag or a misfired transaction can halt a production line, that assumption is a liability, not a valid starting point.
This piece argues the right question isn’t which agent to choose but whether your systems and governance are ready to support one. A few of the core points:
The piece closes with six questions every technology decision maker should answer before deploying, and a clear sequence for success: Get the agent right, get the system right, get the governance right — in that order. Read the full Analyst Insight →
New Thinking on Agent Maintenance
One question I have been thinking about a lot: What will we do to make agents better after the first revision? By extension, what will those pipelines look like, and will there need to be some sort of trade-off between cost and the level of marginal improvements? (For example, is a 2% improvement in accuracy worth hundreds of thousands of dollars in infrastructure and testing?)
Last week, CoreWeave made some interesting announcements that touch on this area, where we are starting to see the idea of autonomous and continuous improvement for agents. I think of this using a variation on the common intern analogy: A new agent is like an intern, and you initially need to teach them everything and provide the right guardrails for them to operate. But at some point, that intern needs to become a junior employee and, instead of needing constant curation, you expect them to learn on the job. Plus you would prefer that anyway, since you’d like to know what a fresh set of eyes could find or break. It’s a pretty innovative idea, and it’s something that I am finding really refreshing with CoreWeave and Weights & Biases right now. Stay tuned, because I am finalizing a deep-dive session with them soon. Meanwhile, here’s the Coreweave blog post if you want to learn more.
Models, ‘Effort,’ and the Markdown-versus-HTML Debate
The last couple of weeks at Anthropic have been interesting. The company made a huge new investment announcement and released a new model, which is being sold as something of an “incremental improvement,” plus there was some good old developer chatter.
I’ll leave the investment talk to others, but in the Opus 4.8 release I did see one very interesting item. In the release notes there was a comment about Opus 4.8’s ability to have Claude Code scale to manage hundreds of sub-agents. Personally, I believe that scalable orchestration will end up being a massive competitive threat to almost any tech services firm, and outsourcing in particular. Part of the reason companies outsource processes — or whole departments — is to hand off the complexity to another company that is supposed to be more skilled and experienced at complex project and process management. And what we are seeing so far is that AI orchestration does not scale well. It is either human-controlled and deterministic (making it brittle) or left to the model or agent, which chews up way too many tokens. So, to see Claude take that on is quite interesting, given that to scale AI we will need very smart orchestration. Oh, and one more thing: 4.87 also introduces the concept of “effort” into the mix. I want to puzzle over that some more as well.
Also in the Claude universe, there was a back-and-forth among developers about what is the better input format for Claude Code. Should we stick with .md or, as some on the Claude team suggest, move to HTML, which is much richer and easier for the human in the loop to parse and understand? Well, unsurprisingly, HTML chews up a lot more tokens than markdown, so there’s some disagreement over which is “better.” My guess is that we may see this debate end up as an optimization in future models, where (somehow or other) we will see a model trained to be far more efficient on HTML since it will provide higher-fidelity inputs, which should lead to better outcomes. We shall see.
NEW ANALYST INSIGHT: Mission-Critical ERP Needs Mission-Critical Agents
Enterprise enthusiasm for AI agents is real and accelerating — but most early deployments rest on a dangerous assumption that the IT systems beneath the agents can stay exactly as they are. In mission-critical ERP environments, where a missed compliance flag or a misfired transaction can halt a production line, that assumption is a liability, not a valid starting point.
This piece argues the right question isn’t which agent to choose but whether your systems and governance are ready to support one. A few of the core points:
The piece closes with six questions every technology decision maker should answer before deploying, and a clear sequence for success: Get the agent right, get the system right, get the governance right — in that order. Read the full Analyst Insight →
New Thinking on Agent Maintenance
One question I have been thinking about a lot: What will we do to make agents better after the first revision? By extension, what will those pipelines look like, and will there need to be some sort of trade-off between cost and the level of marginal improvements? (For example, is a 2% improvement in accuracy worth hundreds of thousands of dollars in infrastructure and testing?)
Last week, CoreWeave made some interesting announcements that touch on this area, where we are starting to see the idea of autonomous and continuous improvement for agents. I think of this using a variation on the common intern analogy: A new agent is like an intern, and you initially need to teach them everything and provide the right guardrails for them to operate. But at some point, that intern needs to become a junior employee and, instead of needing constant curation, you expect them to learn on the job. Plus you would prefer that anyway, since you’d like to know what a fresh set of eyes could find or break. It’s a pretty innovative idea, and it’s something that I am finding really refreshing with CoreWeave and Weights & Biases right now. Stay tuned, because I am finalizing a deep-dive session with them soon. Meanwhile, here’s the Coreweave blog post if you want to learn more.
Snowflake and Databricks both have their big shows in the next few weeks, and you can already feel the jockeying before anyone takes the stage. Snowflake didn’t even wait for its keynote, announcing plans to buy Natoma the week before its Summit conference kicks off. Natoma governs how AI agents connect to enterprise systems and what they’re allowed to do once they’re in by verifying the agent, enforcing its permissions, and keeping a record of every action it takes. What I like is the direction Snowflake is taking, taking the governed-data promise it has sold for years up a level, from governing the data itself to governing the agents acting on it. Snowflake buying this capability instead of building it tells you that the platforms now treat agent governance as too strategic, and too hard, to ship on their own.
Databricks made its own play the same week. It taught its Unity Catalog to govern data sitting on engines Databricks doesn’t own, then reached straight into a rival’s catalog to manage what lives there. Snowflake is pulling governance up toward the agents, while Databricks stretches it sideways across everyone else’s engines. What neither keynote will say out loud is that governing an agent where your data lives only holds until that agent reaches into the next platform over — and most real agents do, touching Snowflake, Databricks, and a handful of SaaS apps in a single workflow. Govern only your own turf and you end up with neat islands of control while the agent roams free in the water between them. So sit through both shows with one question in your back pocket: What governs an agent the second it leaves your platform? Neither company has fully earned that answer yet. That’s exactly why they’ve turned governance into the thing this rivalry is fought over.
It’s hard not to be in awe of Dell Technologies’ most recent quarterly earnings. For the sake of this drill-down, I’m going to focus on its Infrastructure Solutions Group (ISG) results and the signals I’m pulling from them.
First, let’s look at the numbers. It’s clear that the company has become the key player in the enterprise AI infrastructure buildout. ISG revenue reached a record $29 billion, up 181% year-over-year, while ISG operating income grew 206% to about $3.1 billion. Perhaps most notable, Dell maintained operating margins of roughly 10.5%. This is despite the rapid expansion of AI infrastructure deployments — and shows that the company is delivering a strong bottom line to complement its stunning top-line revenue.
The AI infrastructure numbers themselves are almost incredible. Dell reported $24.4 billion in AI orders during the quarter and recognized $16.1 billion in AI server revenue, while AI backlog expanded to a record $51.3 billion. This shows that demand will continue to outpace supply for the foreseeable future. On top of this, management raised its fiscal 2027 AI server revenue outlook to $60 billion, up from prior expectations of $50 billion.
This isn’t just about servers, either. Beyond its AI server boom, Dell also highlighted strong momentum across traditional infrastructure categories, including storage and unstructured data platforms. Again, to me, this is signaling that enterprise AI deployments are increasingly driving broader infrastructure modernization efforts. This is a topic we’ve talked about considerably: Modernization for AI is a datacenter-wide play.
I think the broader takeaway from all this is that Dell appears exceptionally well-positioned for the next phase of enterprise AI adoption. The market is expanding from hyperscalers to enterprises, sovereign environments, service providers, and neoclouds. And Dell’s strength lies in its ability to deliver the entire infrastructure stack. As organizations move beyond experimentation and toward production AI environments, the ability to deliver complete infrastructure solutions becomes critical. The challenge for enterprises is no longer about buying GPUs. It’s building and operating AI-capable infrastructure at scale.
I think what’s even more strategic, though, is Dell’s growing emphasis on software, orchestration, and management. Infrastructure wins create revenue. Management platforms create stickiness.
Dell’s automation capabilities, infrastructure management software, and emerging AI Data Platform strategy position the company to extend its role deeper into the operational control plane of enterprise AI environments. As enterprise AI adoption gains momentum, the vendors that control deployment, lifecycle management, data movement, governance, and operational automation will be well positioned to establish longer-term customer relationships.
The infrastructure side keeps pushing one idea lately, which is that agentic AI should run somewhere you control rather than wherever the public cloud puts it. VAST Data put up the strongest version of this last week when it went live as the data layer underneath Mistral’s new NVIDIA-powered AI factories in Europe. Sovereign AI gets built regularly now, but hardly any single player owns every layer it takes, so it only ships through partnerships. This one is a clean example, with NVIDIA bringing the compute, Mistral the models, and VAST holding and feeding the data — and all of it staying inside the region. For a regulated European buyer, this is a sovereignty story already in production, the kind you can point at and actually copy. One flagship deployment isn’t proof that the architecture generalizes, however, and I want to see VAST under the next three AI factories before I call it a pattern. Even so, it tells you where VAST wants to sit, making its case as the foundation the whole factory runs on rather than a capacity box bolted on the side.
Nutanix takes that control to the agents themselves. It shipped Agent Gateway last week, with governance and cost management for AI agents wired right into the platform its customers already run. I wrote about it when it landed, and the appeal is pretty clear. If you’re already a Nutanix shop sorting out agentic AI, you get to keep that work on the foundation that’s been treating you well, without bolting on a new vendor or splitting governance across systems that don’t talk to each other. It’s worth being honest that the same convenience deepens the vendor lock-in, since governance this easy to adopt is also governance you won’t want to unwind later. Even so, agent sprawl and surprise token bills are the two reasons I keep hearing for stalled agent projects, and putting the controls where your infrastructure already sits removes real friction for getting started. Expect more infrastructure incumbents to make this pitch. As agents move into production, the company already running your stack has the easiest claim on governing what those agents do.
I spent a day at Strategy’s Mini World NYC show last week. A lot of data players are staking a claim to the context layer right now, and Strategy has put one more flag in the ground. It addresses the fundamental issue that arises when you put a model on top of enterprise data with no semantic layer in between, and every agent guesses its own answer, right down to what a basic number like “revenue” even means. Strategy generates the SQL inside that layer instead of leaving it to whichever model is running on a given day, so the answer comes back the same no matter which agent asks. Thirty years of enterprise customers have hardened that engine, and it shows. The catch is that this only holds while someone keeps the semantic model current. So the debate over what (for example) “revenue” means lands on whoever maintains that model, which is at least one clear place to settle it.
On the pipeline side, Astronomer has Otto, its Airflow-expert agent, taking on the translation work that makes migrations miserable. Moving pipelines onto Airflow means rewriting old workflow code, untangling dependencies, and fixing the steps that don’t port cleanly — the kind of slog that routinely drags a migration out for months. Otto takes the first pass at all of it, and because it leans on how Airflow actually runs across Astronomer’s customer base rather than generic model guesswork, the conversions tend to hold up. Then Strategy hardens the answer an agent hands back. With this process, Astronomer takes a lot of the pain out of moving the data reliably in the first place. Neither of the processes I’m describing is glamorous, but this is incredibly valuable plumbing. The reliability and the cost of agentic AI get settled down there, in the semantic layer and the pipeline, not up in the model everyone tends to stare at. If I were running a data team right now, that’s where I’d pressure-test my own readiness, long before the demo that wowed me last week.
Nordic Semiconductor — AI-Assisted Development
Last week, Nordic Semiconductor released MCP servers that give AI assistants such as GitHub Copilot, Cursor, and Claude Code authoritative Nordic platform context: nRF Connect SDK documentation, API references, device configurations, and live field data from nRF Cloud. Those AI tools now handle Nordic-specific development work automatically, eliminating the need for vendor-specific AI tooling. These MCP servers are available now at nordicsemi.com.
Traditionally, edge SoCs ship with proprietary tools tied to vendor-specific IDEs, SDKs, and system architectures. Nordic recognized that modern AI assistants are already capable of autonomous system coding — provided they have full context for the target platform.
Nordic’s MCP servers extend across much of the IoT device lifecycle: device configurations for board bring-up, SDK documentation and API references during active development, and live field data for debugging devices. Connecting all three stages in a single AI conversation is what makes the approach closed-loop — the developer doesn’t lose context when moving from bring-up to prototype to product.
My take: The deeper strategic point is about changing embedded system diversity from a barrier to an enabler. Edge silicon is diverse by design. Different applications demand different power profiles, connectivity options, function blocks, and acceleration architectures. No single chip wins everywhere. That necessary diversity has historically been the developer’s burden to manage — a tax on complexity and a barrier to product variety. With Nordic’s MCP approach, AI assistants handle more of the platform-specific adaptations. The diversity stays in the silicon and becomes less visible to the developer. Done right across the industry, this turns diversity from a barrier into a strategic advantage for customers, who get to choose the best chip for each product without paying excessive diversity taxes.
I’ve been arguing for years that edge semiconductor vendors must deliver more than chips and a GitHub library. The “more” now includes a queryable representation of the platform in the form of MCP servers that AI assistants can directly consume. I expect other semiconductor vendors to follow this same pattern.
Just one week after announcing the Anderon quantum foundry partnership with the U.S. Department of Commerce, IBM has doubled down on its quantum ambitions with a landmark financial commitment. On May 28, 2026, IBM announced a bold plan to invest more than $10 billion in quantum computing over the next five years, formally disclosed through an SEC filing. IBM’s stock jumped by more than 4%.
This announcement follows the recent letter of intent between IBM and the U.S. Department of Commerce to establish a pure-play foundry, named Anderon. The plan is to establish Anderon as a standalone company headquartered in Albany, New York, where IBM has operated semiconductor research and fabrication facilities for more than two decades. Rather than retrofitting an existing standalone fab, IBM is bringing its mature 300-millimeter semiconductor wafer manufacturing capabilities at Albany to bear on quantum processor fabrication, which has operated largely at research scale. IBM’s shift to this 300 mm format at Albany in late 2025 has already produced the Nighthawk and Loon quantum processors, both of which are now in active production.
IBM plans to be the company to build a fault-tolerant quantum computer (FTQC) by 2029. The computer in question has been designated as “Starling” on IBM’s quantum roadmap for some time and is projected to run up to 100 million quantum gates on 200 logical qubits.
The funds committed by IBM are intended to support R&D, capital projects, ecosystem partnerships, manufacturing scale-ups, and mergers and acquisitions as the company pursues its long-term quantum goals. Following IonQ’s lead in quantum M&A — and deviating from its own tradition — IBM may accelerate its roadmap by acquiring external quantum talent and technology instead of relying solely on in-house development. In this plan, Anderon will provide the industrial infrastructure needed to scale quantum computing from laboratory-scale experiments into high-volume production reality. That won’t just be for IBM, either, but also for client companies in the U.S. and beyond. Although it will take years to see the effects of this massive infrastructure move, it will likely play a major role in implementing IBM’s long-term quantum strategy beyond its native superconducting modality.
Viewed from the angle of the Commerce Department’s financial commitment, the Trump administration is aggressively pushing U.S. leadership in AI and quantum computing and other technologies to ensure the U.S. maintains its lead over China.
With $10 billion committed, a named quantum machine on the roadmap, a foundry under development, and substantial U.S. government backing behind it, IBM is making it clear that the quantum era is no longer a distant ambition — it is now a scheduled delivery.
In the lead-up to Computex, Intel announced its newest platform, the Arc G3. The Arc G3 is a derivative of the Panther Lake chip family that the company had already teased at CES 2026, but about which it had given very few details until now. The new Arc G3 is squarely targeted at the handheld gaming segment, but still leverages CPU cores built on the cutting Intel 18A process node, along with Xe3 GPU cores. The Arc G3 Extreme also preserves the Panther Lake Xe3 12-core integrated GPU configuration known as Arc B390, which has been lauded for its performance and power efficiency. We are expecting to see a plethora of Arc G3-based handhelds at Computex, and should see some real competition in a space where AMD has effectively held a monopoly for years since the introduction of the Valve Steam Deck.
Also in the lead-up to Computex, Qualcomm announced its new Snapdragon C processor, which is an addition to its Snapdragon X lineup. The Snapdragon C is the company’s effort to support $300 Windows-based PCs to help combat some of the current challenges with PC pricing and memory supply. Qualcomm has been fairly light on details so far, but we expect there will be some systems at Computex that will feature this chip. While I don’t think the $300 PCs are necessarily there to compete with the likes of Apple’s MacBook Neo, I do think the Snapdragon C is a way to offer decent performance and great battery life to a segment that has really not experienced either. I also believe that if Qualcomm executes this successfully with its OEM partners, it could capture significant share, albeit probably at fairly low margins.
T-Mobile is deepening its 5G presence in golf by entering into a new multi-year USGA partnership, making it the official 5G network provider for the U.S. Open, the U.S. Women’s Open, and additional championships. Among other things, this partnership powers the USGA’s first mobile Rules Review system. It also supports event connectivity and fan experiences. I’ve watched this progression up close. At the PGA Championship last year, I saw in real time how T-Mobile’s 5G and advanced network solutions supported not just fan-facing experiences but also critical workflows in areas including operations, content, and on-site production.
T-Mobile is following a clear path in sports: It started with traditional sponsorship, then moved to sophisticated 5G showcase environments, and now it’s making deployments where 5G, network slicing, and event-specific services are embedded into the operational stack. Sponsorships established the brand and created access to premium venues, and the next phase used major golf championships as advanced test-beds. These events validated performance, density, and new fan experiences under real pressure. Now, with the USGA, those same capabilities are becoming part of the core operations of championship events. They’re used for rules decisions, ticketing, payments, production, and on-course engagement. These are now repeatable, productized patterns that look more like scalable solutions. I look forward to watching how this shift can support enterprise-scale growth for T-Mobile across venues, cities, and vertical industries.
Unless otherwise noted, our analysts will be attending the following events in person.
August events coming soon.
September events coming soon.
December events coming soon.
Founder, CEO and Chief Analyst | + posts
Patrick Moorhead is the founder, CEO, and chief analyst of Moor Insights & Strategy. His big-picture view of technology is grounded in more than 20 years as an executive leading strategy, product management, product marketing, and corporate marketing functions at NCR, AT&T, Compaq, and AMD. He has shared his expertise in areas from silicon to infrastructure to enterprise SaaS and everything in-between in thousands of national broadcast appearances (CNBC, Yahoo Finance), articles (Forbes, CIO), research-based analyses, and podcast episodes. Today, he has 100+ CXO-level advisory clients and is often ranked the #1 technology industry analyst by ARInsights.
Paul Smith-Goodson is the Moor Insights & Strategy Vice President and Principal Analyst for quantum computing and artificial intelligence. His early interest in quantum began while working on a joint AT&T and Bell Labs project and, during 360 overviews of Murray Hill advanced projects, Peter Shor provided an overview of his ground-breaking research in quantum error correction.
Jason Andersen is vice president and principal analyst covering application development platforms, technologies, and services. Jason brings over 25 years of experience in product management, product marketing, corporate strategy, sales, and business development at Red Hat, IBM, and Stratus to his work for MI&S and its advisory clients. Working both in the field and in the headquarters of some of the most innovative technology companies, Jason has a wealth of experience in building great products and driving their adoption across a broad spectrum of industries and use cases.
Senior Analyst-in-Residence | + posts
Bill Curtis is the Moor Insights & Strategy Analyst in Residence for large-scale Internet of Things systems. Bill helps enterprises design distributed solutions that integrate the full end-to-end IoT stack from real-world devices to analytics.
Matt Kimball is a Moor Insights & Strategy senior datacenter analyst covering servers and storage. Matt’s 25 plus years of real-world experience in high tech spans from hardware to software as a product manager, product marketer, engineer and enterprise IT practitioner. This experience has led to a firm conviction that the success of an offering lies, of course, in a profitable, unique and targeted offering, but most importantly in the ability to position and communicate it effectively to the target audience.
Mel Brue is vice president and principal analyst covering modern work and financial services. Mel has more than 25 years of real tech industry experience in marketing, business development, and communications across various disciplines, both in-house and at agencies, with companies ranging from start-ups to global brands. She has built a unique specialty working in technology and highly regulated spaces, such as mobile payments and finance, gaming, automotive, wine and spirits, and mobile content, ensuring initiatives address the needs of customers, employees, lobbyists and legislators, as well as shareholders.
Mike Leone is a principal analyst at Moor Insights & Strategy covering data platforms and analytics, data infrastructure and storage, and data governance and enterprise data strategy. He brings 15 years of analyst experience from his work at Enterprise Strategy Group, where he rose to practice director for data management, analytics, and AI. Mike's work is grounded in a strong technical and strategic foundation, including early roles in software and hardware engineering.
Anshel Sag is Moor Insights & Strategy’s in-house millennial with over 18 years of experience in the IT industry. Anshel has had extensive experience working with consumers and enterprises while interfacing with both B2B and B2C relationships, gaining empathy and understanding of what users really want. Some of his earliest experience goes back as far as his childhood when he started PC gaming at the ripe of old age of 5, building his first PC at 11, and learning his first programming languages at 13.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。