2026 First Half Review
Dear Friends and Colleagues,
This half, our thesis crystallized into a single organizing idea — Throughput Technologies — and our portfolio delivered real, verifiable milestones that validate it. The headline venture numbers are extraordinary but misleading: capital is concentrating at the frontier model layer, which is precisely why we believe disciplined early-stage investing in the coordination layer above it has never been better positioned. We are also watching two dynamics most observers haven't fully priced in: the likelihood of an AI disillusionment cycle before the productivity plateau arrives, and the probability that the energy constraint accelerates efficiency innovation rather than simply inflating costs — both of which, we think, run in favor of where we have spent two years building.
Our Thesis, Sharpened
At Z2Sixty Ventures, we have always believed that great outcomes are born not only from what we do, but from how we think. When we founded the firm, we set out to back visionary founders building global businesses and to deliver outsized returns for our partners. Over the past two years, as we have deployed capital and pressure-tested our thesis against a fast-moving AI cycle, our conviction has crystallized into a single organizing idea we call Throughput Technologies: the coordination layer between artificial intelligence, energy, and human workflows. This letter shares that framework, our progress in the first half of the year, and how we read the market opportunity.
We invest across three pillars of the coordination layer:
Energy Orchestration — the systems that align compute and power with demand under AI's energy constraint.
Data Sovereignty Tools — the infrastructure that lets institutions own, govern, move, and unlock their data, turning tacit knowledge and institutional memory into an accessible asset.
Human-System Bridging — the interfaces and workflows where AI systems augment human judgment rather than replace it, capturing near-term value while full autonomy arrives unevenly.
This is not a departure from how we have always invested. It is a clearer articulation of it. We take an exit view even at the earliest stage, because it informs our conviction on long-term return potential and the nature of the markets a company is entering or creating. We have come to call this discipline our Endgame Framework: we reverse-engineer what a landmark outcome requires and stress-test the ownership and exit math against our return threshold before we invest. From there, we work backwards with our founders — identifying the milestones, hires, product decisions, and market positioning required to actually reach that outcome. That matters more in a market now re-rating towards realized outcomes over paper markups, where the discipline of underwriting to an exit shapes which companies are worth backing in the first place.
Markets Update: A Record Half, and What It Actually Tells Us
The headline numbers for the first half of 2026 are extraordinary. Total US venture funding reached over $411 billion, surpassing all of 2025 ($328 billion). Yet the headline conceals the more important story. Two companies — OpenAI and Anthropic — absorbed more than 53 percent of that capital. AI captured the overwhelming majority of venture dollars, per PitchBook data. Beneath the two behemoths, the market has taken on a barbell shape: enormous late-stage rounds in a handful of frontier names at one end, and disciplined early-stage investing at the other, with a hollowed-out middle in between.
When the marginal dollar flows to the model layer at a scale no emerging fund can match, the returns available to a disciplined early-stage investor migrate to the layer that makes those models usable. That is precisely the coordination layer we have organized ourselves around.
The exit environment reinforces the point, though with important nuance. The first half was solid — 17 $1B+ IPOs and healthy M&A activity year over year — but the durability of that reopening now depends heavily on Anthropic's timing. SpaceX's shares have swung sharply in both directions since its June listing - a roughly 50% drawdown from post-IPO highs followed by a recovery back above the offer price on strong earnings. This underscores how volatile the reception has been for mega-cap tech listings even when the underlying business is performing well, and OpenAI has already signaled a 2027 IPO window as it holds out for a higher valuation. On the M&A side, the strategic activity is the more telling signal: Marvell Technology's $3.25 billion acquisition of Celestial AI to break the AI memory wall through Celestial's optical-interconnect system is precisely the buy-over-build pattern we expect to persist, because incumbents cannot develop coordination layer capability at the speed the market now demands.
Two dynamics we are watching that most observers have not fully priced in. First, AI adoption is likely to hit a trough of disillusionment before it reaches the productivity plateau — the gap between what frontier models promise and what enterprises can actually deploy at scale is real and widening. The companies that survive that trough will be those with genuine contracts and embedded workflows, not demo-ware. Our coordination layer thesis is, in part, a hedge against that cycle: the infrastructure that makes AI usable is not discretionary spend, and it does not disappear when sentiment corrects.
Second, AI's energy constraint is widely read as a crisis. We read it differently: acute, unavoidable pain is a forcing function that compresses sales cycles for the companies solving it. The orchestration layer does not wait for the market to come around — it is pulled in by necessity. That is a tailwind, not a headwind, and it is one reason our energy orchestration conviction has strengthened this half rather than softened.
Portfolio Update
The first half was a period of meaningful, verifiable progress across our companies. In keeping with our practice, the updates below draw only on publicly available developments. We continue to support our founders actively through a still-demanding fundraising environment, where our hands-on work — shoring up business models, sharpening go-to-market, and preparing teams for a more selective capital market — matters most.
A few developments we are proud to share:
DevRev (Data Sovereignty Tools). DevRev is an AI-native platform that unifies customer support and product development around a proprietary knowledge graph — operationalizing AI on top of an institution's own interlinked data rather than sending that data to a shared model. The company grew its team past 900 people this half, launched Computer, its AI teammate product built on that knowledge graph, and expanded its partner ecosystem with additions like Capgemini and Accenture as it scaled EMEA sales coverage. It is an enterprise-grade business executing on the data-sovereignty thesis at scale.
Floto (Human-System Bridging). Floto brings AI-powered design feedback — audits, synthetic-persona testing, flow analysis, and design-to-production comparison — directly into the designer's canvas. Its late May release added design system awareness, detecting mismatches between design intent and what actually ships to production. That is the bridging layer in practice: keeping human teams and AI systems aligned inside the tools people already use, without asking anyone to change their workflow.
Ion Storage Systems (Energy Orchestration). In March, Ion became the first U.S. solid-state battery company to pass customer performance qualification for its Cornerstone cell, following sample shipments to industrial, consumer electronics, and automotive customers. The company expects to begin production at its expanded Beltsville, Maryland facility this year, and in June announced a $2 million venture investment from TEDCO alongside the appointment of a new chief financial officer. It is a textbook example of a capital-efficient hardware company moving from the lab to commercial validation.
ReJoule (Energy Orchestration). ReJoule continued to advance its rapid battery diagnostics platform, which uses electrochemical impedance spectroscopy to assess battery state-of-health in minutes rather than hours. The company also navigated real policy headwinds this half: a federal clean-energy grant was among the more than 320 projects affected by the Department of Energy's funding clawback late last year. ReJoule's response — leaning into strategic partners and private capital, and emphasizing the chemistry-agnostic, bankable value of diagnostics — is exactly the kind of adaptability we look for in the founders we back.
Building the Firm
Z2Sixty itself has evolved meaningfully since we last wrote at length about the firm. What began as a solo general partner effort is now a partnership. Zach Debelak brings twenty five years of product leadership across SaaS, FinTech, and HealthTech; our third partner co-founded a NASDAQ-listed energy technology company and anchors our energy orchestration underwriting with operator depth. We are Chicago Booth classmates who have known and tested each other's judgment for nearly two decades, and our presence across Atlanta, Chicago, and San Francisco gives the fund a genuinely national sourcing footprint.
We have also been active in the broader institutional and technical ecosystem. This summer, Hernando joined a panel on venture capital and family office allocation at the Opal Family Office & Private Wealth Forum in Newport, sharing perspectives on manager selection, portfolio construction, and where we see the coordination layer creating value. Elsewhere across the first half, we participated in Deep Tech Week New York, spent time with the hardware and manufacturing community at mHub in Chicago, and joined the energy conversations at Georgia Tech Energy Day — gatherings that reflect where we see the Throughput Technologies actually being built.
Earned Conviction: What We Updated This Half
We have written before that discipline drives success. Part of that discipline is being honest — with ourselves and with you — about where the evidence has moved us. Conviction that never updates is not conviction; it is attachment.
This half, the case for backing hardware and capital expenditure intensive startups has moved from thesis to conviction. Deep tech and hard-tech investing — categories venture capital has historically shied away from — now attract roughly half of all US venture funding, and one in four unicorns minted in 2026 has come from robotics, defense, and aerospace alone. That momentum is real, but it is concentrated at the top and it is not the whole story. Outside the mega-round dynamics of data-center buildouts and frontier robotics, we see a quieter and more investable pattern: hardware companies whose defensibility comes from combining physical technology with software, supply-chain design, and manufacturing know-how. That combination is difficult to build, difficult to copy, and — as our early support of Ion Storage Systems and ReJoule reflects — increasingly rewarded by strategic acquirers and public buyers alike. This is not a departure from Throughput Technologies; it is the physical layer of it — the energy systems, orchestration hardware, and industrial infrastructure that make the coordination we invest in actually run.
Looking Ahead
We enter the second half of 2026 with conviction and discipline in equal measure. We will be watching whether the reopened exit window pulls mid-cap infrastructure names behind the frontier IPOs, whether AI's energy constraint forces orchestration further toward the center of the market, and whether buy-over-build continues to bring acquisition interest to earlier-stage capability — each of which runs in favor of the coordination layer we have spent two years building around. This remains one of the most selective fundraising environments in a decade, particularly for new managers, and we believe the answer to a selective market is not to broaden our aim but to sharpen it.
As always, we welcome your questions on our strategy, our portfolio, or the broader venture landscape, and we are grateful for your continued interest in what we are building.
With appreciation,
Hernando Bunuan
Founder & Managing Partner
Z2Sixty Ventures
Zachary Debelak
General Partner
Z2Sixty Ventures