The way enterprises use technology is entering a new chapter.

‍

At BGV, we call this shift Enterprise 5.0: the move from software as a tool to software as a worker. AI is moving

beyond copilots that help people complete tasks. Increasingly, AI systems can research, reason, plan and execute

work. We believe some of the most important companies of the next decade will become the systems of action

inside enterprises, owning critical workflows rather than simply helping users manage them.

‍

This builds naturally on what BGV has been doing for more than two decades: investing at the frontier of

enterprise technology. Today, we invest globally in AI-native companies at the Seed and Series A stages, with a

particular interest in businesses tackling high-friction, labor-constrained industries where AI can deliver clear

and measurable value.

‍

Within Enterprise 5.0, there are four areas where we are especially excited to meet more founders. These are not

the boundaries of our investing. They are simply the themes where our thinking has been most active recently.

‍

Energy: Powering the Intelligence Economy

‍

AI may be digital, but the infrastructure supporting it is very physical. As demand for compute grows, power is

becoming a critical constraint on AI infrastructure. Access to energy increasingly determines where and how

quickly compute can be deployed. We see energy not simply as another vertical, but as a foundational layer of

the AI economy.

‍

Our primary interest is in the broader AI compute infrastructure stack and the technologies required to deliver

reliable compute at scale. We are looking at opportunities around power and interconnection, cooling,

infrastructure management, workload orchestration, GPU scheduling and energy optimization. We are also

interested in distributed and behind-the-meter generation, storage and technologies that can make existing grid

capacity more useful.

‍

As data centers evolve from passive consumers into more active grid participants, we see an emerging

opportunity for software that can match AI workloads with available power in real time. Longer term, we are also

following energy finance, digital power markets and industrial electrification as AI reshapes how energy is

produced, financed, managed and consumed.

‍

Physical AI: Bringing Intelligence Into the Real World

‍

The first major wave of generative AI has largely lived on screens. The next will increasingly interact with the

physical world.

‍

Robot foundation models and world models are converging with lower-cost sensors, actuators and edge

computing. Together, these shifts are making robots and autonomous systems more capable, adaptable and

economically practical, moving them from task-specific automation toward more general-purpose capabilities.

‍

We are particularly interested in capital-efficient Physical AI companies with strong technology and data moats.

This includes embodied and multimodal intelligence, specialized models trained on spatial, kinetic or force data,

perception and sensor fusion, and software infrastructure that helps physical systems learn and operate reliably

in the real world.

‍

We believe the data engine behind these systems will be especially important. Teleoperation, simulation and

world-model pipelines can turn real-world operation into training signals, allowing deployed systems to

continuously improve.

‍

On the application side, we are especially interested in manufacturing, warehouses and logistics, supply chain

and construction. These are areas where automation can address repetitive, difficult or hazardous work and

where customers can see a clear return on investment.

‍

Healthcare: Toward Autonomous Care Operations

‍

Healthcare combines enormous amounts of data, complex workflows and highly skilled people spending too

much of their time on administrative work. The first wave of healthcare AI showed that models can draft notes,

process information and assist with individual tasks. We believe the next wave will increasingly own entire

workflows, with clinicians supervising rather than executing every step.

‍

We are particularly interested in AI that moves beyond point solutions to transform how care is delivered and

paid for. This includes agentic systems for prior authorization, claims, revenue cycle and member navigation. We

are also interested in clinical agents that monitor patients and coordinate care between visits, AI-native

approaches to value-based care, and tools that compress clinical trial and evidence-generation timelines.

‍

As these systems become more autonomous, the infrastructure around them becomes increasingly important.

We are looking at evaluation, monitoring, clinical-grade guardrails, interoperability and data infrastructure that

can turn outcomes, clinician corrections and real-world feedback into continuous learning.

‍

Across these areas, we are looking for companies that can build durable advantages through proprietary data,

deep workflow integration, distribution and trust. As autonomy increases, the ability to demonstrate safety,

auditability and measurable outcomes will become increasingly important.

‍

Commerce: From Search to Execution

‍

For the past two decades, e-commerce has largely been built around humans searching, comparing and clicking.

We believe the next wave will increasingly be built around agents that act.

‍

Agentic commerce moves AI from recommending a product to discovering, evaluating and executing

transactions on behalf of consumers and businesses. This shift is already underway as AI platforms, retailers and

payment networks build infrastructure for agent-led transactions.

‍

We are looking for startups building across this new commerce stack. This includes product and catalog

intelligence, pricing and compliance, AI-native merchandising and discovery, as well as agents transforming

procurement, demand forecasting, inventory, warehouse operations, logistics, delivery and returns.

‍

As agents begin transacting with other agents, we also see a new trust layer emerging around agent identity,

delegated authorization, spending controls and fraud prevention.

‍

The opportunity is not simply to add AI to today's e-commerce software. It is to rethink what commerce

infrastructure looks like when software can make and execute decisions itself.

‍

What Connects These Themes

‍

Energy, Physical AI, healthcare and commerce may look like very different markets. To us, they are connected by

the same underlying shift:

‍

AI is moving from generating answers to executing work.

That is the heart of Enterprise 5.0.

‍

We want to back founders building AI-native companies that own important workflows, solve real enterprise

problems and create measurable economic value. We are especially interested in markets where workflows are

complex, labor is constrained and the ROI from automation is clear.

‍

This is why we are particularly excited about vertical AI. We believe much of the value will be captured by

companies that deeply understand their industries and workflows, and combine proprietary data with

organizational and industry context to build increasingly intelligent and autonomous systems.

‍

These themes are not the full map of what we invest in. We continue to back exceptional founders across

enterprise technology, and we expect the categories themselves to keep evolving. Some of the most interesting

companies may sit at the intersection of several of them, or create a category we have not yet named.

‍

If you are building an AI-native startup in one of these areas, or believe your company belongs in the Enterprise

5.0 conversation, we would love to hear from you. Reach out to the BGV team. We are always happy to meet

founders building what comes next.