Jeremiah Owyang

Last week Massachusetts Institute of Technology published a sobering report on the stark reality divide between the hyped exuberance of AI investments and disappointing outcomes of 95% of GenAI enterprise pilots yielding zero returns. While this led to a cautionary wave in public markets, this scenario echoes the trajectory of cloud platforms in the early- to mid-2010s where where initial hype and trials eventually paved the way for widespread adoption over the next decade. 🔹 A key insight in the MIT report was that opting to "buy" AI solutions often proves more fruitful than attempting to "build" them from scratch. DIY approaches were successful only 1/3rd of the time vs. buying off-the-shelf AI tools resulted in a success rate of 67%. Enterprises don't need to re-invent the wheel; instead leverage your key strengths and core differentiators to build on top of best-of-breed vendors. 🔹 The report also highlighted the main reason why AI pilots fail - flawed enterprise integration. The "learning gap" for both tools and talent is real in a world where technology moves as the speed of light. The tools in itself can be great only when they are appropriately integrated into workflows and the rest of the tech stack, as well as training and re-skilling the workforce to use the tools. We are still in very early innings of the enterprise AI shift where AI is poised to become the cornerstone of software stacks, similar to what cloud is today. Enterprise adoption takes time and often cycles through growing pains as what we're experiencing today. The forthcoming decade promises transformative changes, paving the way for the emergence of groundbreaking enterprise AI companies that will shape the future landscape.

82

17 Comments

Hardware enabled software will be the most defensible companies in the AI ERA. Hardware-enabled SaaS turns sensors and devices into proprietary data engines, then converts that data into sticky, high value software with elite retention. The top companies check the boxes that Jeff Horing at Insight Partners highlighted on his recent interview on Invest Like The Best: Big ROI, Big ASP, Fast Time-to-Value. How it looks in the real world 1. Axon ($56B Market Cap) - Law Enforcement Tech What they sell: cameras and software that help police record what happened and finish paperwork faster. The Result: $2B+ Revenue and 124% Net Revenue Retention. Why it wins: devices create the evidence, Evidence.com and AI draft the reports, and multi-year bundles keep customers for years. 2. Badger Meter ($5.38B Market Cap) - Water Tech What they sell: smart water meters and software that help cities find leaks and bill accurately. The Result: $826.6M revenue in 2024. Why it wins: meters stream usage data, BEACON flags leaks and losses, and long replacement cycles plus data history raise switching costs. 3. Samsara ($22.62B Market Cap) - Safety for the Physical World -What they sell: trackers, dash cams, and software that help fleets and job sites run safely and on time. -The Result: $1.35B ARR growing mid-30s percent year over year. Gross logo retention cited at 98% plus. Net Revenue Retention 115-120%. -Why it wins: devices instrument vehicles and equipment, the Connected Operations Cloud runs daily ops, and multi-product expansion fuels retention. Why now Component costs keep falling, connectivity covers the edge, and AI is turning raw signals into decisions in seconds. This enables counter positioning. Give away or subsidize the device, price software on outcomes, and let the data flywheel deepen the moat. My simple HES playbook 1. Start with a wedge that instruments a high value, high frequency decision. 2. Remove capex friction with free or subsidized hardware that undercuts incumbents dramatically 3. Close the loop so alerts trigger real work orders in the system of record or complete the action that operators had to do historically by leveraging AI (e.g. report writing) 4. Compound moats as every deployed device grows your proprietary dataset that trains your models. 5. Sell outcomes - increased revenue, direct labor savings, insurance savings, loss reduction, etc. I'm going to share a deep dive in these businesses and my thesis in coming weeks. Let me know any other companies you admire that you think I should study!

149

26 Comments

💡 Speed is above all for building AI infrastructure Meta is quietly building one of the world’s largest AI training clusters in Ohio (Prometheus). Inspired by xAI’s unprecedented time-to-market, Meta is embracing a #datacenter design that prioritizes speed above all else. Zuck threw his entire datacenter playbook into the trash and is now building multi-billion-dollar GPU clusters in “Tents”! 🔎 Self-build campus. 🔍 Leasing from third parties. (they pre-leased more capacity second half of 2024 than any hyperscaler) 🔎 AI-optimized designs. 🔍 Multi-datacenter-campus training. 🔎 On-site, behind-the-meter natural gas generation. (when the local power grid couldn’t keep up, Meta went full Elon mode. With help from Williams, they’re building two 200MW on-site natural gas plants. ) 🔍 Connect all these sites with ultra-high-bandwidth networks all on one backend network powered by Arista 7808 Switches with Broadcom Jericho and Ramon ASICs. While all eyes are on the high-profile Stargate datacenter in Abilene, Meta has been planning a response for over a year and making tremendous progress. The Louisiana cluster is set to be the world’s largest individual campus by the end of 2027, with over 1.5GW of IT power in phase 1. (Sources tell us this is internally named Hyperion.) And, Meta has many other datacenters under construction and ramping. Read more: https://lnkd.in/gJJgaWVT #AIDataCenter #AIInfrastructure #LargestTrainingCluster #VerticallyIntegratedDataCenter

57

6 Comments

On-Prem is the New Cloud: Why Enterprises are Reversing Stance in the Age of AI Remember when cloud was the battleground for innovation, agility, and scale? Everyone was tracking the % of on-prem compute moving to the cloud and it was growing every year? Well now there’s a reversal. The reality today? On-prem is the new cloud. Here’s why. AI’s Dark Side: Legal Risk and Data Governance Headaches Enterprises are excited about AI, but also uneasy as vendors like OpenAI face intensifying legal pressure. In a high-stakes lawsuit, a judge ordered OpenAI to preserve all ChatGPT chats (even deleted ones), overriding its 30-day deletion policy. Imagine that: you delete sensitive conversations, but they’re still sucked into legal limbo. The precedent is clear: AI prompts and outputs are discoverable records, demanding integration into enterprise ESI (Electronic Stored Information) policies. For large organizations handling sensitive data, the risk is simple: you can’t afford your AI tools to turn into legal liabilities. Especially when global regulations (GDPR, etc.) and internal governance collide with unpredictably broad discovery orders. Sovereignty, Control, Trust Many enterprises still ban ChatGPT or impose strict rules on usage. Even when allowed, token limits are throttled. Enterprises are cautiously letting AI in (as they should). Enterprises need to have control and trust. SAP is leaning into this with its Sovereign Cloud On-Site, deployable on customer premises with SAP-managed infrastructure. It is built to deliver data, operational, technical, and legal sovereignty in one package. Doubt it will be the last. Scaling Pains Talking to founders, three reasons stand out why cloud isn’t scaling with AI: 1/ The rate at which AWS/GCP lets you provision compute and the compute it wants you to provision both falter as AI and AI driven developers deploy a lot more code more frequently than before. So sometimes it’s not about data governance, but actually about performance and functionality. 2/ Shadow AI is rampant—employees paste sensitive info into models despite restrictions—creating demand for better permissioning tools. 3/ Data leakage remains a live risk, like the August 2025 Grok incident where hundreds of thousands of private chats became publicly accessible and indexed. One thing is for sure - founders selling into the enterprise are scaling faster if they offer the ability to run on-prem. The Future is Hybrid Don’t get me wrong - cloud is not going anywhere. But neither is on prem. A few years agoI thought we were headed for a full cloud shift, but now I think the future is truly hybrid.

33

4 Comments

🚀 The AI-Native Revolution: Why Re-engineering SaaS Won't Cut It in the Enterprise 5.0 Era After decades of SaaS dominance, we're witnessing a fundamental shift that's reshaping how businesses operate and compete. This isn't just another tech upgrade—it's a paradigm change from traditional software to AI-native architectures. 💡 The Dawn of Enterprise 5.0 We're entering Enterprise 5.0, where human ingenuity combines with intelligent machines to create autonomous workflows. Unlike previous eras, employees become orchestrators of intelligent workflows rather than manual processors of information. 🏗️ Architecture is Destiny The difference between AI-native and traditional SaaS isn't in surface features—it's in architectural DNA: Traditional SaaS: Static workflows, bolt-on AI features, fragmented data AI-Native:Intelligence embedded from the ground up, continuous learning, unified data flows Adding AI features to legacy SaaS is like attaching a jet engine to a horse-drawn carriage. You might go faster, but you haven't built an airplane. ❌ Why SaaS Re-engineering Fails Three critical failure modes: 1. Fragmented Data Access - AI features only see siloed data, missing the full picture 2. Broken Learning Loops - Insights don't flow across systems, limiting compound intelligence 3. Task vs Process Automation - Bolt-on solutions automate tasks, not entire workflows 🎯 The Real Challenge Building AI-native solutions faces unique hurdles: • 85% of AI projects fail due to data quality issues • Only 12% of organizations have AI-ready data • 67% of employees show low-to-moderate AI acceptance • Requires entirely different tech stacks and organizational transformation 📈 The Market is Moving VCs have invested $2+ billion in agentic AI startups in the past two years. 25% of companies using generative AI will launch agentic AI pilots in 2025, growing to 50% by 2027. 🔮 The Path Forward Success requires: ✅ Full-stack AI rearchitecture ✅ Human-centric design that amplifies capability ✅ Different economic models (higher initial investment, lower marginal costs) ✅ Continuous learning architectures The Bottom Line:Just as budgets moved from TV to digital and software migrated to cloud, this transition will be swift and decisive. Companies retrofitting AI onto legacy architectures will be outpaced by AI-native competitors built for this new era. The future belongs to those who understand that we're not just adding intelligence to existing software—we're fundamentally reimagining how software and services work. At BGV, we're backing founders building these human-centric, AI-native solutions that combine the best of human ingenuity with intelligent machines. This is Enterprise 5.0, and architecture is destiny. #AI #Enterprise #Startups #VentureCapital #AINative #Enterprise50 #BGV #FutureOfWork --- What's your take?

43

6 Comments

Anthropic passes OpenAI to become #1 in enterprise AI! Enterprise use is exploding ($3.5B to $8.4B in 6 mos), with coding the first killer app. Menlo is a leader in enterprise AI research - read our whole report here: https://lnkd.in/gXXHeUBA

110

6 Comments

Voice AI has come a long way, and it is finally starting to sound natural enough for real conversations at scale. This week, Claremont Colleges founder Lily Clifford shares how she left Stanford’s PhD program to start Rime Labs and build speech models that can handle millions of real-world calls every day. She talks about what it takes to personalize voice at scale and the starting point for building voice AI that actually works. It’s a Claremont Colleges world out there. 👇 https://lnkd.in/gKGZjEzB

38

5 Comments

The next frontier of AI isn't just about choosing the right model. It's about building moats through implementation. Across the Emergence Capital portfolio, companies like Bland and Federato are scaling faster by embedding Forward-Deployed Engineers. These engineers bridge the gap between AI and reality, ensuring real results in production. Read more from my colleagues Gordon and Kyle on why FDEs are the key to unlocking AI's full value. https://lnkd.in/gcsRRGX2

29

3 Comments

ICYMI. AI is driving the cost of intelligence to zero. This innovation is going to spark rapid and massive transformations in tough industries: eg drug discovery, advanced mnfr, energy, mobility, defense, etc. https://lnkd.in/gRTinSxp

56

11 Comments

Fascinating report published by MIT and reported on in Fortune. The report highlights that only 5% of AI pilot projects successfully scale, with most failures occurring in internal development rather than vendor-led initiatives. Integration emerges as the primary barrier to scaling AI projects. The current Silicon Valley approach emphasizes sales and marketing, which benefits startups but is less impactful for large enterprises where a minimum of 1% revenue lift is required to get any C-suite attention. In these organizations, the real challenge lies in integrating AI across diverse constituencies, hierarchies, and applications. This complexity is evident in typical SAP implementations, where consulting and integration costs far exceed the application itself. I saw this first hand building Pantero -- a semantic broker. Time and time again, enterprise architects favored expediency and speed via cheap XML integration hacks using XSLT -- I know so old!! instead of building and leveraging metadata to ensure extensability and re-use, especially since systems change regularly. The report highlights that most AI solutions today focus on sales and marketing, which is not where enterprises see the greatest impact. Instead, significant value comes from simplifying processes across multiple systems and organizational layers. This requires a strong service orientation, exemplified by companies like Palantir, which build solutions around integrated services. But AI lacks an inherent orchestration capability. The analogy I like to draw is that the chat interface is the equivalent of the basic line prompt which so many marveled at back in the early Bill Gates days but real impact is going to require the equivalent of an agentic IDE. Hence. it is no surprise that AI agents deployed externally for customer service are logical initial use cases due to minimal integration needs. However, scaling AI across multiple systems with complex semantics, workflows, and logic demands coordinated orchestration of multiple agents—adding a layer of complexity beyond simple chat interfaces. But the notion of orchestration raises the larger question of how agentic a service can be if it has to be orchestrated. By definition autonomy demands less overt decision making but orchestration seems to collide with that approach. Where we are investing and see great success is where agentic ai monitors stateless message requests across organizations instead of inside them. Or where there is no dependency on multiple systems to complete a task. #agenticai #ai #aifail #aisucceed #aiorchestration #orchestration ping us at www.xrcventures.com/apply if you have a solution or are working on something. The link to connect with us is in the first comment. https://lnkd.in/evsvaM3s

27

2 Comments