“Most investors still treat AI and crypto as separate worlds. The real opportunity li

“Today, most investors still view AI and crypto as separate worlds. The real opportunity lies in the infrastructure layer that connects them: protocols that allow agents to discover counterparties, negotiate, validate and settle transactions programmatically,” said Ran Levitzky, general partner at Magenta Venture Partners.

When asked which sector or trend is currently being ignored and which he believes represents the most undervalued opportunity for the coming year, Levitzky highlighted the AI-Crypto convergence. “As AI agents become autonomous economic actors, they must be able to cryptographically identify themselves, hold value, and transact without human intermediaries or centralized payment rails,” he said.

“It’s about building the transactional fabric for agent-to-agent commerce. The winners in this category will effectively become the connective tissue of the machine economy.”

You can read the entire interview below.

Fund name: Magenta Venture Partners

Total assets under management: 100 million dollars

Partners/Managers: Ori Israely, Ran Levitzky, Daisuke Makino, Ari Dotan

Notable holding companies: Workiz, Onebeat, Findings, Sensos, QG, Monogoto

Notable releases: Autotalks (acquired by Qualcomm), Valens Semiconductor

The Valuation Jump: Beyond the market correction, what is the most critical metric (e.g. EBITDA, NRR) that will determine premium valuations in 2026?

Net revenue retention will be the primary metric, especially above around 110%. It captures what the market currently rewards most: real product value, organic expansion within existing accounts, and efficient growth rather than purchased growth through burn. Companies exceeding 120% NRR will continue to benefit from premium multiples, since capitalization comes from their existing customers rather than incremental CAC.

The important change is that the market no longer pays for “AI wrappers”. AI alone is not a divide. Premium valuations will accrue to companies with domain defensibility: proprietary data loops, deep integration of hard-to-eliminate workflows, and clear capital efficiency.

The Agent Leap: As we move from “co-pilots” to autonomous “agents,” which specific vertical will be the first to fully trust AI with independent decision-making?

Logistics and supply chain appear to be the first to adopt it on a large scale. It is already a data-rich, automation-driven environment where decisions are frequent, structured and measurable.

We are now getting to the point where an AI “operations manager” can reroute shipments, adjust inventory, and respond to exceptions in real time autonomously. (Magenta Portfolio Company) Sensos, for example, deploys agents who monitor global maritime flows and manage disruptions. (Magenta Portfolio Company) Onebeat uses autonomous decision-making to continuously optimize retail inventory rather than relying on rigid planning cycles.

On the service side, field service operations are also evolving rapidly. (Magenta Portfolio Company) Workiz’s AI assistant “Jessica” already answers calls, understands intent, and independently books jobs for home service companies. Confidence grows because the return on investment is immediate: fewer missed calls and busier schedules.

These types of environments, where outcomes are measurable and risks are non-lethal, will welcome autonomous agents first.

The Sovereign Leap: Have the geopolitical lessons of recent years pushed Israeli startups to build independent and “sovereign” technology stacks to reduce their dependence on global platforms?

There is certainly a greater awareness of resilience and dependence, but there is no mass abandonment of hyperscalers. Israeli startups still rely heavily on AWS, Google Cloud, and Azure due to their convenience and dominance of the ecosystem.

What has changed is the mentality rather than the choice of platform. Founders are increasingly designing with optionality in mind: multi-cloud deployments to reduce lock-in, greater use of open source tools instead of entirely proprietary stacks, and better attention to data portability.

The lesson of recent years is not to “go it alone,” but rather to “have a plan B for critical infrastructure.”

The Leap in Efficiency: In the age of AI-driven hyperproductivity, has the traditional correlation between “workforce growth” and “business success” been broken for good?

In many cases, this correlation is now structurally broken. AI-native companies can achieve tens or even hundreds of millions of dollars in ARR with teams that are a fraction of the historical norm.

We see startups improving their efficiency by 15-25x, with revenue per employee levels ranging from $500,000 to over $1 million, which is no longer unusual. Customer support, quality assurance, and much of routine operations are becoming increasingly automated, while product teams move faster with AI-driven development.

The new signal is not how many people you hire, but how much production and revenue a small team can produce. Success is decoupled from headcount and recoupled to capital efficiency and production.