About
Experienced Product and GTM SaaS executive adept at leading product strategy…
Articles by Michael
Activity
6K followers
-
Michael Ni shared thisAmong a slew of announcements, Elastic’s quarter was solid: revenue up 15% to $478M, sales-led subscription revenue up 18%, cRPO up 21% and RPO up 27%. The market is looking for signs of improving momentum (e.g., growing pipe) as it sorts out which vendors AI creates real operating leverage for. OK, Elastic’s numbers are improving, but the more interesting question is whether the product footprint (and signs of consolidation ... or increased ability to compose solutions) is starting to line up with where enterprise AI actually needs infrastructure. AI agents operating real systems need more than a model. They need access to current operational context. That's telemetry, security signals, search, incident history, and state ... things Elastic largely sits across. I'm watching if Elastic can turn its assets into a shared operational context layer for both people and agents. That would make the company look less like a search or observability vendor and more like part of a runtime infrastructure for AI-driven operations. For more details, check out Mike Wheatley's article https://lnkd.in/gkdF3HHXAfter posting a solid earnings beat, Elastic's stock bounces higher in extended trading - SiliconANGLEAfter posting a solid earnings beat, Elastic's stock bounces higher in extended trading - SiliconANGLE
-
Michael Ni shared thisIf you are in the New York area, join us as we bring data and AI leaders together to drill into pragmatic lessons on operationalizing AI.Michael Ni shared this🚨 AI isn’t waiting. Is your organization ready? On September 24 in New York City, executives, data leaders, and AI pioneers will come together at AI Forum 2026 to move beyond the hype and focus on what’s next. From AI strategy and governance to generative AI and enterprise adoption, we’ll tackle the questions leaders need to be asking now: → What’s real—and what’s still hype? → How do you turn AI experimentation into measurable business value? → What does responsible AI leadership look like? → How do organizations build the foundation to scale AI with confidence? Real strategies. Actionable insights. Executive perspectives. 📍 The Harvard Club, New York City 📅 September 24, 2026 🔗 Save your spot: https://bit.ly/4xB10P9 #AI #AIF2026 #ArtificialIntelligence #GenerativeAI #AILeadership #DigitalStrategy
-
Michael Ni posted this10 takeaways from recent @DisrupTVShow episodes https://bit.ly/4gCeJO4 @ConstellationR Research's DisrupTV is a weekly wisdom machine that churns out lessons from a bevy of thought leaders. Here's a look at the takeaways from recent episodes.
-
Michael Ni shared thisCongrats to the DuckLabs team on their acquisition by Amazon Web Services (AWS). My read: this is less about replacing centralized analytics and more about influencing/gap-filling where analytics can run. For those wondering, Redshift and Athena still matter. DuckDB complements those engines with a complementary layer for lightweight, embedded analytics where moving data or spinning up a warehouse is overkill. That fits a broader shift toward workload-centric architecture: run the job where it makes the most sense based on cost, latency, governance, and where the data already lives, instead of defaulting to a centralized database... a line of thinking that benefits AWS. Importantly, by not acquiring the DuckDB open-source project itself means DuckDB, DuckLake, and Quack stay open under the Foundation. At the same time, AWS gets the core team and much tighter roadmap alignment with S3 and its analytics stack. MyPOV: For data + AI leaders, the implications: • No need to force any workload through the same platform • Define clear rules/governance for when work runs in the warehouse, against object storage, inside an app, or closer to the edge • Data platform teams will need stronger orchestration and workload-placement capabilities to span distributed estate • Put one team in charge of workload placement, cost, latency, and runtime standards across data and AI as choice spans data, ops, and dev teams The bigger shift: the control point in analytics (and increasingly decisioning) is increasingly not just the database. It is deciding what should run where ... and under what rules. Good piece from Mike Wheatley at SiliconANGLE on why this deal matters beyond DuckDB itself. https://lnkd.in/gwvVvaVEAWS buys DuckLabs to bring DuckDB's embeddable analytics to more enterprises - SiliconANGLEAWS buys DuckLabs to bring DuckDB's embeddable analytics to more enterprises - SiliconANGLE
-
Michael Ni reposted thisMichael Ni reposted thisWhen it comes to enterprise AI, context, trust, and sound judgment are becoming just as important as the technology itself. In the latest Constellation newsletter edition, our analyst insights explore what it takes to move #AI forward, from evolving enterprise context and AI-ready #data to governance, #agenticAI, and the role of human judgment. What's inside: • Round two of the 2026 Q3 Constellation ShortLists™ • A closer look at what comes after RAG for enterprise context • What orgs need to build trusted foundations for agentic AI • How governance and human judgment will shape AI’s next phase • An upcoming webinar on securing and governing data for agentic AI that you won't want to miss Let’s get into it 👇The Enterprise AI Shift: Context, Trust & What Comes NextThe Enterprise AI Shift: Context, Trust & What Comes Next
-
Michael Ni shared thisCost per token is the wrong metric. Time to think: Return on Intelligence. Stripe’s acquisition of OpenRouter is an important signal. Stripe built its business by sitting in the transaction flow. OpenRouter puts it in another rapidly growing flow: AI consumption. First, the basics being reported: this is a $7.5BLN bet on owning a piece of the token flow... one that fits Stripe well: sit in a high-volume transaction path, meter usage, manage scale economics, abstract complexity, and make it easy for developers to consume. Second, what matters for data + AI leaders is that intelligence isn’t a commodity in the same way payments are. Models may commoditize, but they won’t become interchangeable. This matters because enterprises will deliberately dial the level of intelligence to the task. At the end of the day, why send a $10 problem to the most powerful model available if a cheaper model reliably delivers the required outcome? That makes model routing a strategic control point as AI consumption scales ... a point not lost on Boomi, Snowflake, and even NVIDIA, all of whom released their own model routing capabilities as gateways, FinOps, and orchestration start to converge. Companies controlling the model-allocation layer are capturing an increasingly important position in the AI value chain. Read more on the acquisition on Constellation Research, Inc. Insights: https://lnkd.in/e7x_fiyE
-
Michael Ni shared thisShould your AI agent automatically become you? As copilots shift to agents, that assumption starts to break down. In my conversation with Waqas Ahmed, VP of AI Engineering at OpenText, we discussed the difference between an assistant acting on behalf of a user and an autonomous agent operating continuously across systems, potentially with other agents... and therein lies an open architectural question. Enterprise security was largely designed around humans: establish identity, assign permissions, grant access. But as agents become independent actors, simply inheriting a user's permissions may give them far more authority than the task requires. The emerging challenge isn't just giving agents identities. It's defining their operating boundaries. That means what they can access, what they can do, and in what context. As Waqas points out in this clip, that changes how we need to think about trust as agents gain autonomy. You can check out the full conversation with Waqas here https://lnkd.in/exQvRriS
-
Michael Ni shared thisFinOps for AI is becoming much more operational. Snowflake’s launch of dynamic model routing is another signal: don’t send every task to the biggest model. Dial the level of intelligence to match the value of the job. What’s different? CDAOs and CIOs shouldn’t treat this as just cost optimization. It’s the early rise of runtime economics for AI. At runtime, enterprises will increasingly decide: Is this model good enough? Does it meet the latency and quality required? Is the decision worth the cost? Does spending more intelligence actually improve the outcome? MyPOV: FinOps metrics will move from cost per token to cost per decision, even as model routing moves ROI discipline to become executable and not just observable ... even as tokenmaxxing moves to value. Read more from Larry Dignan on the rise of model routing https://buff.ly/xw60lY0Snowflake to add dynamic model routing to Cortex AI GatewaySnowflake to add dynamic model routing to Cortex AI Gateway
-
Michael Ni shared thisThe AI theme this year: moving from agentic demos to safely scaling agents in production across real enterprise systems. That is where the questions get harder and more practical: • Can the agent access live operational data? • Does it understand the business or customer context? • Whose identity and permissions apply? • What policy governs the action? • Can you trace ... and increasingly replay what happened afterward? Excited to drill into these questions in a fireside chat with two practitioners, Rex Washburn (Chief Data Architect, CDW) and Arif Rajwani (Snr Solution Architect, Denodo), on what Data and AI leaders need to get right before agents can safely act across ERP, CRM, analytics, and custom applications. Just did the run-through - we're cutting through the AI governance theory into production failure points, lakehouse boundaries, logical data access, active context, and runtime governance. Join us live, Wednesday, August 26 at 10:00 AM PDT And come with your questions! Register here: https://buff.ly/9yGtK5w
Michael Ni replied to a comment
5h
Nicholas Cort with context-washing already a buzz word, next up is showcasing use and impact. Thanks for the comment.
Michael Ni commented on a post
4d
Steven Dickens helped us kickoff our doggie appreciation! Here's my grand lady (17 years old) enjoying being an urbanite here in Taipei this week versus our standard tree lined Palo Alto walk
-
Michael Ni liked thisMichael Ni liked thisAnthropic's launch of Claude Fable 5.1 and Mythos 5.1 is a win for enterprise customers who pushed back on zero data retention and pricing. Anthropic's launch addressed data retention, price and then performance. Reading between the lines, this launch is an interesting tell that Anthropic lacks the pricing power you'd think. https://lnkd.in/eAdRXDb9Anthropic's Claude Fable 5.1, Mythos 5.1 launch addresses price and data retention concernsAnthropic's Claude Fable 5.1, Mythos 5.1 launch addresses price and data retention concerns
-
Michael Ni liked thisSharing our great collaboration story from the upcoming September issue of The CEO Magazine. Bespin has been working closely with Neology over the past three years to modernize the technology foundation of aging state highway systems. Neology provides leadership in modern tolling technologies, data processing, and digital payments, while Bespin contributes our expertise in cloud, AI, and cybersecurity. We are incredibly proud of this strong partnership and the amazing results we continue to achieve together!Michael Ni liked thisGrowth means little if you can’t deliver on it. In the latest feature from The CEO Magazine Global, Neology, Inc. CEO Bradley H. Feldmann, NACD.DC shares how disciplined execution, technology, and customer outcomes have helped transform the company into a rapidly growing force in transportation technology. As a strategic partner to Neology, Inc., Bespin Global US is proud to support that journey — helping bring together cloud modernization, managed operations, cybersecurity, Amazon Connect, and AI to build the reliability, scalability, and innovation needed to grow with confidence. As our CEO, Sunny Kim, puts it, Neology, Inc. is exactly the kind of strategic customer relationship Bespin is built to support. Because transformation isn’t just about adopting new technology, it’s about turning complex technology decisions into measurable outcomes. Read the full feature in The CEO Magazine Global: https://lnkd.in/dGuJMrXF #BespinGlobal #CloudTransformation #AI #Cybersecurity #CloudModernization
-
Michael Ni liked thisMichael Ni liked thisHuzzah! Our boys head off to school... and my wife and I can get back to work. Happy Back To School Day ✏️📚🚌
-
Michael Ni liked thisMichael Ni liked thisI missed National Dog Day, so here’s Maple getting ready to party this weekend!
-
Michael Ni liked thisMichael Ni liked thisPeople were asking me about my walk-on music for my Dreamforce presentations. Here you go. #DF26
Experience & Education
Languages
-
Chinese
-
Recommendations received
25 people have recommended Michael
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content