Monthly US Tech Investment Report

Palm Drive Capital
March 2026

Executive Summary

Palm Drive Lens

These trends reinforce Palm Drive’s positioning across three core theses: (1) AI applications and vertical software are where enterprise value accrues—our portfolio in Deep Instinct, Fullseam, Cranston, and others captures the shift from horizontal AI to purpose-built solutions; (2) the data layer is the true differentiator as model capabilities converge—Datacurve and Traverse sit at this critical inflection; and (3) Physical AI and robotics (via Medra, Boom Supersonic, and Impulse Labs) represent the next frontier as enterprises move from chat interfaces to embodied, autonomous systems.

This winter, Palm Drive assigned one team member to evaluate companies with repeat founders—he ended up talking to 105. One standout from the pipeline: a repeat founder building an AI-native workflow platform for a regulated vertical, who closed a $12M Series A with strong unit economics and a clear path to $10M ARR by year-end—exactly the kind of capital-efficient, vertical AI play that defines our sweet spot.


AI Landscape

Macro Trends: Record Capital & Regulatory Clarity

The first quarter of 2026 marked an unprecedented inflection point for AI investment. Global venture funding reached $300 billion across 6,000 startups, with AI capturing 80% of all capital ($242 billion). Four of the five largest VC rounds ever recorded closed in Q1 alone: OpenAI ($122B at $852B valuation), Anthropic ($30B), xAI ($20B), and Waymo ($16B) collectively accounted for 65% of global venture deployment.

On the regulatory front, the White House released its National Policy Framework for Artificial Intelligence on March 20, outlining seven pillars including IP protection, innovation enablement, and federal preemption of fragmented state laws. While not binding legislation, the framework signals a coordinated federal push to create uniform rules—a critical relief for companies building at scale.

Infrastructure Layer: The $1T Race for Compute

AI infrastructure has become the defining capital allocation story. Amazon, Google, Meta, and Microsoft are deploying nearly $700 billion on infrastructure in 2026 alone. Nvidia’s Blackwell Ultra (GB300) platform is tracking to ship 60,000 racks in 2026, a doubling from 2025 demand. However, the real bottleneck is not just GPUs—memory, interconnects, and power delivery now define what is possible.

Inversion Semiconductor’s miniaturized laser-driven particle accelerators for EUV lithography target a 15x speed improvement in chip fabrication efficiency—addressing a structural bottleneck that constrains the entire AI supply chain.

NVIDIA GTC 2026 Takeaways: Jensen Huang unveiled the Vera Rubin platform (10x perf/watt over Grace Blackwell) and previewed the 2028 Feynman architecture. Purchase order projections hit $1T through 2027. Agentic AI platforms (OpenClaw, NemoClaw) and physical AI partnerships (Uber autonomous fleet in 28 cities, BYD/Hyundai Level 4 vehicles) dominated. DLSS 5 brings full neural frame generation. Key signal: infrastructure spend is accelerating, not plateauing.

Data Layer: Synthetic Data Goes Operational

The conversation around AI data has shifted from theoretical to operational. NVIDIA announced the Physical AI Data Factory Blueprint on March 16, an open reference architecture automating how training data is generated, augmented, and evaluated. Gartner forecasts that by end-2026, 75% of enterprises will use generative AI to create synthetic customer data.

This trend validates the positioning of Datacurve, which provides curated, high-quality code as training data for LLM developers, and Traverse, which captures real human judgment and behavioral data to create training environments for frontier models. Both are now critical infrastructure plays as model quality increasingly depends on data quality over parameter count.

Applications Layer: AI Agents Move from Lab to Enterprise

March 2026 marked the moment AI agents shifted from experimental to embedded. Anthropic’s Model Context Protocol surpassed 97 million installs, signaling its transition to foundational infrastructure. Alibaba launched Wukong, an enterprise AI platform for managing multi-agent workflows.

Enterprise adoption is accelerating across accounting (Cranston, Fullseam), sales enablement (Kepler, Throxy), security (Deep Instinct, Zania), and insurance (Panta). The market is moving from “what can AI do?” to “how do we deploy and manage AI at scale?”—a shift that favors purpose-built vertical solutions and workflow automation.


Hardware & Robotics

Physical Robotics: The $1.2B March

March 2026 saw over $1.2 billion deployed into robotics in a single week. The market bifurcated sharply: autonomous mobility (robotaxis and self-driving trucks) captured over $18 billion of $19.3 billion total, with companies like May Mobility (autonomous shuttle fleets) and Moove (fleet financing for ride-hail in emerging markets) demonstrating that the real value is in deployment, not just hardware. Industrial humanoids and general-purpose robotics emerged as the next capital frontier.

Palm Drive’s Medra (autonomous wet lab protocols for drug discovery) and AutoPallet Robotics (warehouse swarm automation) are well-positioned in adjacent niches where enterprise productivity gains are measurable and high-margin.

Transportation & Energy Hardware

Electric vehicles remain in transition. 32 new EV models launched in the US in 2026, with prices falling as battery costs dropped 20% in 2024. Electrified vehicle research jumped to 23.8% of shopper activity in mid-March—the highest of 2026.

Boom Supersonic announced Symphony engine testing at Colorado Air & Space Port in 2026, with FAA certification targeted for 2029–2030. United Airlines CEO Scott Kirby publicly pegged Boom’s odds at “50-50” in late March—a comment that reflects the genuine technical risk in supersonic certification but understates Boom’s progress: Symphony is the first purpose-built supersonic engine in decades, and the Colorado tests are a concrete milestone, not a concept paper.

Meanwhile, Impulse Labs is gaining traction with next-gen electric home appliances, including a 36-inch induction cooktop launching March 2026.


Deep Tech

Space: Industrialization of Orbit

The space tech sector is transitioning from exploration to commercialization. Orbital life extension services, optical communications, and lunar logistics are attracting significant capital.

Palm Drive portfolio company AstroForge is executing on its asteroid mining vision. With $55.5M in funding and founders from Virgin Orbit and SpaceX/NASA, AstroForge’s Vestri mission—scheduled for 2026 launch—will attempt the first private landing on a celestial body outside the planetary gravity well.

Semiconductors: Domestic Fab Race Accelerates

Tesla and SpaceX jointly announced “Terafab,” a $25 billion chip fabrication facility in Austin designed to consolidate design, lithography, fab, memory, packaging, and test in-house. This signals a structural shift: mega-cap tech is now willing to build fabs to secure AI chip supply.

The EUV lithography bottleneck remains critical. ASML maintains near-monopoly control, but Inversion Semiconductor is attacking this with laser-driven particle accelerators that could deliver 15x faster chip fabrication. Success here is a generational shift in semiconductor capacity.

NVIDIA GTC 2026 Takeaways: Jensen Huang unveiled the Vera Rubin platform (10x perf/watt over Grace Blackwell) and previewed the 2028 Feynman architecture. Purchase order projections hit $1T through 2027. Agentic AI platforms (OpenClaw, NemoClaw) and physical AI partnerships (Uber autonomous fleet in 28 cities, BYD/Hyundai Level 4 vehicles) dominated. DLSS 5 brings full neural frame generation. Key signal: infrastructure spend is accelerating, not plateauing.


Software & Enterprise SaaS

SaaS funding activity surged in March 2026 with notable rounds: Cloaked ($375M Series B, privacy/security), Xbow ($120M, automated penetration testing), Dash0 ($110M Series B, agentic observability). AI integration has become the defining characteristic—over 31% of recently funded B2B startups incorporate AI/ML capabilities.

This surge coexists with growing talk of a “SaaSpocalypse”: the fear that AI agents will collapse entire SaaS categories by automating the workflows those tools were built to support. The reality is more nuanced—AI is reshaping SaaS, not killing it. Commoditized horizontal tools face real displacement risk, but vertical SaaS with deep workflow integration and proprietary data moats is actually strengthening. The winners will be platforms that embed AI to deepen lock-in, not standalone tools that AI can replicate.

Palm Drive’s portfolio in enterprise infrastructure—Carta (cap table/valuations), Rippling (HR/payroll/devices), Cockroach Labs (distributed SQL), and Addepar (wealth management)—continue to benefit from enterprises consolidating and automating core workflows.


US vs China: Competitive Landscape

US-China technology competition intensified in March 2026 on multiple fronts. Congress approved the Chip Security Act on March 26, which embeds tracking technology directly into advanced chips—a fundamental shift in enforcement from export licensing paperwork to physical tracking. On March 19, Super Micro Computer co-founder Yih-Shyan Liaw was arrested for conspiring to export Nvidia chip-equipped servers to China via Taiwan and Southeast Asia intermediaries between 2024–2025.

China responded with a massive new subsidy program for AI and semiconductor development, doubling down on self-sufficiency goals to overcome US export controls. Meanwhile, the US administration is navigating trade negotiations, with chip export controls becoming a political pressure point.

For Palm Drive: our applications and data portfolio stands to benefit regardless of which country leads in frontier models—enterprise software demand is global. Inversion Semiconductor has upside if it can deliver on faster lithography, but our core exposure is to the AI adoption wave through vertical software and data infrastructure.


Portfolio Spotlight

AI — Security

Deep Instinct: Deep Learning for Zero-Day Protection

Deep Instinct continues to gain traction in endpoint protection, with mindshare increasing from 0.7% to 0.9% in the Gartner EPP category. Its preemptive security approach (predicting and preventing threats in <20ms, 750x faster than ransomware) aligns with enterprise demand for autonomous defense.

Hardware — Robotics

Medra: Automating Wet Lab Discovery

With $63M in funding, Medra is automating repetitive wet lab protocols to accelerate drug discovery. As pharma R&D increasingly relies on robotics and AI, Medra’s niche is becoming a critical productivity lever for biotech.

Hardware — Energy

Impulse Labs: Home Electrification Goes Mainstream

Impulse Labs’ 36-inch induction cooktop, launching March 2026, represents the consumer-facing edge of home electrification. As electric grid modernization accelerates, smart home appliances become key infrastructure.


Palm Drive Insight: Founder Archetypes

This winter, Palm Drive assigned one team member to evaluate companies with repeat founders—he ended up talking to 105. Approximately 70% had a repeat founder. The pipeline skewed toward AI/ML applications (~33), defense/aerospace (~18), healthcare/bio (~15), and fintech/payments (~14), with 46% of companies at pre-revenue or under $50K and only 14% above $1M ARR at first call.

Studying this cohort revealed a clear pattern: the conventional signals investors use to evaluate repeat founders—exit count, school pedigree, previous investors on the cap table—are far less predictive than behavioral inputs like why they’re really doing this again, what they talk about when you don’t ask, and whether the customer or the investor comes first. The founders with the best inputs had the strongest traction.

Three Archetypes

Archetype #1: The Domain Dominator

These founders don’t have a sales cycle—they have a trust cycle. They build in the same market they just left, where “prove yourself” is already done. Relationships come first; the product is secondary. The result is 3–4x faster sales cycles in company #2. The risk: the domain they know becomes a ceiling, and they build for customers they have rather than customers they’ve never met.

Archetype #2: The Revenge Arc

The productive ones have moved past the story. They talk about the problem, have nightmares about product quality, and channel resentment into execution. The destructive version is still explaining whose fault it was at the last company—the resentment drives distraction, not focus. The best founders in this archetype have moved past the narrative and back into obsession.

Archetype #3: The Serial Operator

The hungriest ones weren’t the ones with the biggest exits. The $100M operator left because something was unfinished—the last company’s mission isn’t done yet, and they’re building with 10x fewer people. The $10M operator had a clean exit and good returns but “just wanted to build again” with no specific problem pulling them back.

The Rule

Our evaluation framework for repeat founders distills into four criteria. First, domain carry matters more than exit count—the key question is “where did your first customer come from?” Second, traction velocity: is company #2 growing faster than #1 with less capital? Faster growth on less burn is the strongest signal. Third, the exit gap is not a red flag—what matters is why they’re doing this again. “I wasn’t done” beats “just wanted to build again” every time. Fourth, character score: how they respond to hard questions, whether they’re agentic, whether they push to see term sheets rather than wait.


Looking Ahead: April–May Key Watch Items