Our 2025 Outlook

We're at a fascinating crossroads. In the last few months alone, we've watched AI go from a novel concept to a near-ubiquitous layer in just about every new startup. It's hard to ignore how quickly founders have shifted from "Should we use AI?" to "Our entire product and operations are informed by AI — what's next?" Alongside this, we're seeing a fundamental rethinking of how companies operate, hire, and scale. The old playbooks from the past 15 years — shaped by giants like Google, Netflix, or Meta — aren't as relevant in a landscape where smaller teams can achieve world-class results by leveraging ever-more-accessible AI models.

At Antigravity, we're adjusting our lens accordingly. Our core focus remains early-stage investing, but as we look ahead to the next 12 months (and beyond), we see several clear themes emerging:

Shifts on the Horizon

AI Normalization

AI is no longer a mere talking point. It's the default assumption — especially for savvy founders who see it as integral to their business models rather than a gimmicky selling point. That means differentiation can't come from simply saying "we have AI." It has to come from how you build distribution, weave AI into workflows, and deliver tangible value. The result is a higher bar for quality, reliability, and real market traction.

Small Teams, Big Impact

We've never seen such a dramatic compression of what a handful of talented builders can do. Accessible AI libraries, open-source frameworks, and cheap cloud infrastructure empower small teams to achieve what once required entire floors of engineers. This matters for capital requirements and fundraising timelines. It also demands a fresh perspective on how we assess a startup's "scale" at the seed stage — lean doesn't mean inconsequential anymore.

M&A Motives Evolving

The days of pure acqui-hires are fading. Larger companies are less inclined to buy a startup simply for its engineering team when AI coding tools make it cheaper than ever to build software in-house. What acquirers want now are specialized data sets, a proven go-to-market engine, or deeply integrated solutions that can't be trivially cloned. For founders, it means developing genuine moats, not just product demos.

Large-Scale Infrastructure and Public-Private Partnerships

Multi-billion-dollar initiatives — from HPC clusters to major data-center expansions — are on the table. We're seeing governments and big corporates form consortiums around AI, robotics, and simulation platforms. This opens doors for startups that supply specialized software or can plug into vast data streams, potentially gaining rocket-fuel traction if they align with these emerging mega-ecosystems.

Sector-by-Sector Recalibration

Industries like healthcare, education, and logistics are inching toward AI adoption, but it won't happen overnight. It calls for creative approaches to distribution and an understanding that these fields move slower for very real reasons: regulation, entrenched workflows, bureaucratic purchasing cycles. Cracking these markets is possible but demands patience and a willingness to craft custom solutions rather than expecting a self-service signup flow to do all the heavy lifting.

The Three Investment Themes Powering Our Thesis

As we watch these trends play out, our investment lens zeroes in on three core categories that we believe are ripe for both near-term traction and long-term transformation:

1. AI Infrastructure

Why It Matters

AI might be "table stakes," but the infrastructure behind it is anything but trivial. From MLOps and data pipelines to security, governance, and compliance, this "plumbing" is the foundation that keeps AI systems reliable and integrated with enterprise workflows. We see infrastructure companies laying down sticky moats; once they're deeply embedded, customers rarely churn.

What We Look For

• Platforms that handle data orchestration and model management at scale
• Governance and compliance solutions for increasingly strict AI regulations
• Observability and security layers that ensure trust in AI deployments
• Performance optimizations for HPC and advanced compute environments

2. AI Agents

Why It Matters

We're transitioning from AI-as-a-tool to AI-as-an-autonomous (or semi-autonomous) worker. Whether it's an AI SDR automating sales outreach or a multi-agent platform coordinating entire workflows, these solutions promise radical gains in efficiency. We see a near-future where specialized AI agents handle tasks across every vertical, freeing human talent for higher-level creativity.

What We Look For

• Vertical-specific agents in domains like finance, healthcare, or logistics
• Agent orchestration platforms that chain multiple specialized AIs together
• Hybrid "service + software" models that deploy human oversight where needed
• Proprietary data or domain expertise that sets one agent apart from the next

3. Industry Transformation

Why It Matters

Certain sectors — healthcare, education, defense, and beyond — move slowly yet represent massive opportunity. Companies that blend specialized AI platforms with on-site experts can replace consultant-heavy processes that drag on for years. The payoff is huge: once a startup becomes a trusted partner in a high-barrier industry, others in that sector often follow suit.

What We Look For

• Deep domain expertise enabling specialized solutions in regulated or high-stakes fields
• Real traction with pilot customers who have entrenched legacy systems
• Strong "forward-deployed" or "service + software" approach that nails on-the-ground integration
• Credibility-building methods, from compliance certifications to proven ROI metrics

The Modern Startup Organization

AI is more than a product feature. It's reshaping the DNA of the modern startup itself. Rather than defaulting to the old playbooks from tech giants, we see a wave of founders rethinking how they structure teams, build culture, and approach go-to-market. With AI woven into every layer, these companies can rapidly iterate, adapt to market signals in near-real time, and operate with a fraction of the headcount that once seemed mandatory.

This approach demands a first-principles mindset about everything from hiring to how product roadmaps are formed. Is a large sales force necessary, or can an AI-driven approach to lead generation dramatically reduce overhead? Do we need rigid team divisions, or can cross-functional squads pivot faster when AI-automated tasks handle half the workload?

Founders embracing this new model question every legacy assumption. They're not shy about throwing out conventional wisdom if it no longer applies in an era where AI can handle coding, customer support, data analysis, and more. By cultivating a culture of speed, experimentation, and continuous learning, these startups become highly adaptable — arguably a key advantage when technology and market conditions shift so rapidly.

(We’ve written more about this here.)

How We're Refining Our Thesis

Against this backdrop, we're sharpening how we evaluate seed-stage opportunities. Below are some guiding principles that inform our investment decisions:

1. Distribution from Day One

AI alone doesn't guarantee a market. Whether it's enterprise, consumer, or a messy regulated sector, we ask: How does this startup plan to land its first ten (or hundred) paying customers? Founders who nail distribution strategies early — or are at least ready to experiment aggressively — tend to separate themselves from purely tech-driven competitors.

2. Service + Software for Higher Stickiness

We've seen how "mini-Palantir" hybrids can thrive. By wrapping a high-touch service layer around a robust AI engine, you solve immediate, high-value problems — and develop deeper relationships with customers. That intimacy can then evolve into a scalable software platform as the product matures and the client is hooked by real ROI.

3. Capital Efficiency

Small teams can now do extraordinary things. We favor founders who recognize that huge budgets and oversized headcounts aren't always prerequisites for success. Sometimes, controlling burn rate and focusing on profitability or near-profitability early is the most reliable path to building a durable company.

4. Ripe Verticals and Specialized Data

We keep an eye on areas like BPO, healthcare, edtech, and other verticals where large manual workflows can be replaced or augmented by AI. We're cautious about "single bets" in complicated markets without domain expertise, but when a founder truly understands the landscape — and has a plan to navigate sales obstacles — we're all in.

5. Thoughtful AI Partnerships

With major HPC projects spinning up across the globe, forging ties with large cloud providers, data center operators, or specialized HPC resources can be a game-changer. Whether that means discounted GPU credits or integration into a broader ecosystem, these strategic alliances can accelerate time to market and keep overhead down.

Our Evolving Operating Approach

We're not locking ourselves into rigid sprints or narrow targets. Instead, we're committed to staying curious about opportunities that fit our three core themes — whether they're horizontal B2B apps, consumer plays, or some hybrid that defies easy labels. That said, we still uphold a rigorous process to ensure we invest in genuinely high-potential teams:

• Discovery & Diligence: Alongside evaluating the technical, we perform a "distribution check," because the best AI product in the world won't succeed without a viable go-to-market. We also examine how deeply AI is integrated into the startup's operations — does it power just a flashy feature, or is it embedded throughout the organization?
• Post-Investment Collaboration: Our role doesn't end with a check. We help founders refine their "service + software" balance, connect them with domain experts, and share real-world best practices for building a modern AI-driven company from scratch. We believe in cultivating a supportive network where knowledge flows freely, so every portfolio company benefits from each other's breakthroughs.
• Community & Knowledge Sharing: We're building a community of AI-forward operators, CTOs, and domain specialists who understand how to reimagine the entire startup organization — beyond just writing code. Through events and regular meetups, founders can trade notes on hiring, cultural alignment, and adopting AI best practices at every level of the business.

The Road Ahead

We see a future where AI is seamlessly woven into every layer of modern startups — no longer a novel add-on but a core design principle. In some industries, that means adopting advanced infrastructure or orchestrating swarms of specialized agents. In others, it means forging deep partnerships to reshape entire sectors that have resisted change for years.

At Antigravity, our commitment is to back the earliest-stage founders who aren't just building an AI product, but an AI-driven company. That means focusing on the three pillars of AI Infrastructure, AI Agents, and Industry Transformation, all supported by a new approach to organizational design. We believe the companies that master these elements will be the ones redefining the tech landscape — solving real problems at scale, forging sticky relationships with customers, and doing it all with surprising efficiency.

Over the next 12 months, we'll invest with this refined thesis in mind, leaning into founders who share our conviction that the AI era is just getting started. We're excited to learn from and support entrepreneurs who see "normalizing AI" as the starting point, not the finish line. By blending near-term practicality with a bold vision for transforming entire markets, we believe we can collectively shape an era of innovation that outlasts the current hype cycle.

The opportunity is enormous, the bar is higher than ever, and we couldn't be more thrilled to keep building alongside the most daring founders of this new decade.