In partnership with

Where are we in the Trough of Disillusionment?

As we come to the mid point of 2025, AI is a present-day reality for most orgs. Yet, AI stands at a pivotal intersection: it’s powerful enough to require decisive action, but still ambiguous enough to warrant caution.

Interestingly we can use the Gartner Hype Cycle model for clues. It’s a generally helpful model guiding leaders on when emerging technologies will deliver tangible value. When applied to enterprise AI, the message is clear: we’re moving past the “peak of inflated expectations” and slowly entering the “trough of disillusionment.”

The initial surge of excitement was undeniable. ChatGPT shattered adoption records, copilots launched across sectors, and vendors injected AI into every conceivable product. Gartner’s 2024 report highlighted how enterprise AI platforms, foundation models, and AI-driven software engineering soared to the top of the hype curve.

Experimentation is no longer the norm. The F5 2025 State of Application Strategy Report reveals that 96% of enterprises are actively deploying AI models, with half already implementing AI gateways to manage model interactions with infrastructure and applications.

Now however, challenges are surfacing. Budgets have been stretched thin on pilots that failed to scale. Model hallucinations have eroded trust. Infrastructure limitations have slowed progress. What was marketed as seamless has proven to be deeply complex.

On the other hand, we have also seen positive adoption figures. A recent NVIDIA survey report that AI challenges are steadily receding, with 50% fewer firms citing budget constraints and significant drops in data and privacy concerns, signs that enterprise AI is stabilizing beyond the early hype curve.

The trough is where genuine learning takes place. Inflated promises give way to operational realities. Enterprises stop chasing flashy demos and start asking the critical question: where does AI actually deliver value?

Gartner forecasts we are 12–24 months away from the ‘slope of enlightenment’ for most enterprise AI use cases. Best practices will solidify, tool stacks will stabilize, and clear ROI benchmarks will emerge. Some Signs of progress are already visible. With 98% of executives planning to boost AI infrastructure spending in 2025 and over 70% reporting revenue increases from AI, enterprise adoption is ROI-driven.

Success will require a new perspective on AI’s role in the enterprise. AI isn’t just an add-on or a sidebar chatbot. It represents a fundamental shift in how knowledge is managed, decisions are made, and work is executed.

A new enterprise stack is emerging, one engineered for intelligence:

• Unified data layers that provide models with structured, governed context

• DevOps pipelines reimagined for MLOps and continuous model deployment

• Agent-based systems orchestrating tools across business functions

• Governance frameworks focused on model output instead of just data lineage

And it’s not only the technology that must evolve, but the operating model as well. AI adoption is a dynamic capability, adapting with every model update, regulatory shift, and change in computing economics.

So, what’s next for leaders? Are you still chasing headlines, or are you building systems for repeatable, scalable use cases? Are your teams skilled in prompt engineering but unsure about managing organizational change?

The upside is enormous for those who proceed with focus and discipline. Over the next 18–36 months, enterprise AI will reach its “plateau of productivity.” when the advantage shifts from early adopters to integrators. Organizations that don’t just use AI, but build with it.

Tech Triangle. It’s a thing.

On June 9, Stak Ventures and Silicon Valley Impact hosted an inspiring Monday lunch gathering for ecosystem builders, advisors, investors, and startup. Bringing together pure energy and bright smiles in a stunning live/work/play space. (The lunch time old-fashions helped a little 🥴)

If there’s a new Oakland on the rise, bold and ambitious projects like this will be its foundation. Offering a fraction of the costs compared to San Francisco and the South Bay, all just a BART ride away, this development creates a practical, accessible pathway for future founders to enter and thrive in Silicon Valley’s ecosystem.

👯‍♂️🎉 Congratulations Abacus Semiconductor!

On June 11, we attended the grand opening of Abacus Semiconductor’s new HQ in the heart of downtown San Jose. This is a major milestone not only for the Abacus team but for the broader deep tech community in Silicon Valley.

The event drew an impressive crowd, including Mayor Matt Mahan, who led the ribbon-cutting and emphasized the city’s commitment to supporting innovation and the growing AI ecosystem. The strong presence from the City of San Jose signals a renewed energy downtown and a clear vote of confidence in Abacus’s bold vision

Stay Classy Opal Friends

Another exceptional event hosted by our friends at Opal Group. Two days at the Hyatt Embarcadero San Francisco where family offices, investors, and allocators came together for networking and idea exchange.

The room was filled with sharp, sophisticated leaders engaging in strategic conversations, backed by a stellar speaker lineup and a fantastic setting. Proud to see such a successful outcome.

Looking forward to Napa!

//Whats Next

Bay Area Startups Collectively Secured $16.6B in June

$16.6B went to Bay Area startups so far in June, with the bulk of that investment made by Meta into Scale AI. Close to, but not quite an acqui-hire, Meta agreed to invest $14.3B, valuing Scale at more than $29B, and Scale's founder and CEO, Alexandr Wang agreed to move to Meta to run a new AI “superintelligence” group.

M&A: That steady flow of acquisitions that started in January is continuing into June, with twenty-three so far this month. While twenty of those were for undisclosed amounts, the three that did disclose had a total transaction value of just over $2B.

Fast forward to IPOs: With Chime's successful IPO completed – they priced higher than their range at $27, opened even higher and closed their first day at $37.11, a bump of 37% - we're expecting to see more companies queueing up to finally go public. The Chime IPO was a down round – at their inital pricing, their fully diluted market cap was $11.6B, more than 50% down from the $25B valuation they received in 2021. The success of their IPO reflects investors starting to accept down-round IPOs as part of a necessary reset from the frothy 2021-2022 period. More on that from Pitchbook here.

Follow us on LinkedIn to stay on top of what's happening in 2025 in startup fundings, M&A and IPOs, VC fundraising plus new executive hires & investor moves.

Early Stage:

• Mosanna Therapeutics closed a $80M Series A, a clinical-stage biotech company pioneering a novel pharmaceutical approach to treating obstructive sleep apnea (OSA) with an easily administered nasal spray.

• Bedrock Ocean Exploration closed a $25M Series A, a vertically-integrated ocean exploration company developing proprietary robotics and software to quickly and cheaply explore the entirety of the Earth’s oceans.

• Somnee closed a $10M Seed,on a mission to transform sleep and neurological health while fundamentally improving the human experience with the Somnee Smart Sleep headband.

• Volantis closed a $9M Seed, a semiconductor startup building photonically integrated computers for the AI era and the first to scale photonics inside computer chips.

• Aethero closed a $8.4M Seed, building the next generation of space-grade computers and autonomous satellite platforms.

Growth Stage:

• Meter closed a $170M Series C, provides internet infrastructure for businesses plus everything needed to get great internet, networking, Wi-Fi, and cell coverage in a space.

• Linear closed a $82M Series C, helps streamline software projects, sprints, tasks, and bug tracking.

• PostHog closed a $70M Series D, provides product analytics, heatmaps, session recording, web analytics, feature flags, and A/B testing all in an integrated, open source platform.

• Pactum closed a $54M Series C, an AI-based system that helps global companies to automatically offer personalized, commercial negotiations on a massive scale.

• Beewise Technologies closed a $50M Series D, we are the leading global solution provider for pollination services, on a mission to save bees in order to secure the global food supply.

Follow us on LinkedIn to stay on top of what's happening in 2025 in startup fundings, M&A and IPOs, VC fundraising plus new executive hires & investor moves.

“If you know where the data lives—don’t recalculate it.”

Shimon Ben-David, CTO, WEKA

At our AI INFRA SUMMIT, Shimon Ben David didn’t give a talk—he dropped a playbook. In a no-fluff, high-signal session, he pulled back the curtain on what it actually takes to scale retrieval-augmented generation (RAG) and inference in production.

• RAG vs. Fine-Tuning: Shimon clarified what most teams are still wrestling with—how to choose the right approach. His verdict? RAG wins when time-to-insight and model agility matter most.

• Weka Warp & AMG: He introduced the infrastructure magic under WEKA’s hood—architectures purpose-built to smash GPU memory ceilings and accelerate token throughput.

• The New Stack Metrics: KV cache efficiency, vector DB integration, and data locality are no longer “advanced topics.” They’re the new foundations of performant AI.

• Production-Grade Wisdom: His takeaway was deceptively simple but deeply strategic: Don’t rerun what you can retrieve. In the age of LLMs, memory is compute.

Shimon’s session wasn’t a roadmap. It was a reckoning—for every infra team serious about turning inference into impact.

Your Feedback Matters!

Our mission is to provide an insider's view of Silicon Valley's undercurrents – insights often overlooked by mainstream sources. While many newsletters offer broad market overviews, we focus on delivering a unique, in-depth understanding of the local ecosystem. We share behind-the-scenes conversations, introduce key players we meet at events, and offer exclusive insights.

Your feedback is crucial in helping us refine our content and maintain the newsletter's value for you and your fellow readers. We welcome your suggestions on how we can improve our offering. [email protected] 

Logan Lemery
Head of Content // Team Ignite

StartEngine’s $30M Surge — Own a Piece Before June 26

Private markets are having a moment, thanks to companies like StartEngine.

The leading alternative investing platform is helping everyday investors like you access deals once reserved for VCs and insiders, including exposure to private market titans like OpenAI, Databricks, and Perplexity.¹

How’s it going? In Q1 2025, StartEngine pulled off $30M in revenue, its biggest quarter ever (based on unaudited financials).²

But StartEngine isn’t just a middleman. The company earns 20% carried interest on select pre-IPO offerings, unlocking value for shareholders when these deals succeed.³

How can you tap into this diversification play? By investing in StartEngine.

StartEngine has crowdfunded $85M+ to date, and you can join 45K+ shareholders before the company’s current round closes on June 26.

Invest in StartEngine

_{Reg A+ via StartEngine Crowdfunding, Inc. No BD/intermediary involved. Investment is speculative, illiquid & high risk. See OC and Risks on page.}