Innovating Capital Targets AI Infrastructure Investment Wave
As autonomous systems mature, capital shifts toward secure identity and orchestration layers to enable enterprise AI deployment
Artificial intelligence (AI) investment is entering a more disciplined phase, with infrastructure rather than hype now driving capital allocation. Venture capital is moving away from surface-level generative tools and toward identity, security, and orchestration layers that allow enterprises to deploy autonomous systems at scale.
The shift reflects a growing recognition that agentic AI cannot operate inside banks, retailers, or logistics networks without clear audit trails, permission controls, and cross-cloud interoperability. As enterprises experiment with autonomous decision-making, investment is concentrating on the foundational systems that turn experimental AI into operational infrastructure.
“In order for companies to adopt a lot of this and actually use it, there have to be identity layers, security, and compliance controls,” said Anthony Georgiades, a general partner at Innovating Capital, in an interview with TechJournal.uk during the Digital Assets Forum in London on February 5. “If I’m giving an AI agent access to transmit money and make a transaction, it touches a lot of potential identity access.”
Founded in 2017, Innovating Capital is a Greenwich, Connecticut-based venture capital firm with a presence in New York. It has backed roughly 50 technology companies across cybersecurity, fintech, Web3, quantum computing, enterprise infrastructure, and data platforms.
Rather than treating AI as a standalone vertical, Georgiades sees it as a force multiplier that requires foundational upgrades across cloud environments, authentication systems, audit logs, and data governance pipelines.
Enterprises are confronting the operational reality that AI agents cannot simply be plugged into legacy Enterprise Resource Planning (ERP) systems or customer databases without rethinking access permissions and audit trails.
That requirement is reshaping due diligence. Rather than asking how quickly a model can be trained, investors are examining how securely it can be deployed.
Questions around data lineage, permissioning, cross-border compliance, and integration with existing enterprise software stacks are moving to the center of board-level discussions. For Innovating Capital, this shift reinforces the view that durable value will be created by companies solving structural bottlenecks, not by those merely layering AI onto legacy workflows.
Georgiades is openly skeptical of what he describes as an oversupply of shallow AI products flooding the market.
“There’s a lot of what I would call AI slop — a lot of noise, not a lot of real stuff going on,” he said.
The distinction, in his view, is between surface-level generative features and deeply embedded enterprise systems that can withstand regulatory scrutiny and operational stress.
Instead of chasing rapid user growth, Innovating Capital evaluates whether a company strengthens data integrity, improves cross-cloud interoperability, or reinforces identity governance. These layers may be less visible to end users, but they determine whether AI systems can function reliably inside regulated industries.
First Bucket of Gold
Georgiades’ conviction in digital infrastructure stems from personal experience. Before entering venture capital, he was pursuing graduate studies in computer science and robotics, with a focus on computer vision. Exposure to cryptography and early blockchain concepts shifted his trajectory toward decentralized systems.
He made early personal investments in cryptocurrency, encouraged in part by conversations with his then-girlfriend, now wife, who had written an academic paper on pre-ICO (Initial Coin Offering) token models. Those investments generated capital that helped launch Innovating Capital’s first fund, in which Georgiades contributed roughly 10–15% of the total capital.
That early exposure provided both technical literacy and capital flexibility at a time when institutional participation in digital assets remained limited.
“When things were going like this, what did most investors do? They went out, looked at their unrealized portfolio, and raised money from it. We weren’t raising money then. We were selling our assets,” he said.
By selectively exiting high-performing positions during the 2021 crypto surge, Innovating Capital returned gains to investors and preserved liquidity. Rather than expanding aggressively at peak valuations, the firm chose to consolidate and prepare for a correction.
Georgiades said discipline created strategic flexibility. When markets tightened in 2023, Innovating Capital was positioned to deploy capital into companies building long-term infrastructure rather than short-term speculation.
That experience reinforced his belief that technological revolutions unfold over cycles, not quarters.
Infrastructure Over Hype
Innovating Capital’s portfolio reflects a deliberate spread across enabling layers rather than a narrow thematic bet. While AI infrastructure is central, the firm also backs foundational technologies such as quantum computing and blockchain finance. It focuses on core infrastructure across identity, data, finance, computing, and next-generation hardware rather than consumer-facing applications. The portfolio includes:
Tie — an AI-powered identity platform that helps e-commerce brands identify site visitors, enrich first-party data, and convert anonymous traffic into measurable revenue. Resolving identity across devices and sessions enables retailers to reduce their reliance on third-party cookies and build durable, privacy-aware customer datasets.
Pineapple Financial — a Canadian mortgage fintech developing a digital asset treasury strategy and working to bring real-world mortgage data on-chain through standardized, tokenized records. The approach aims to create machine-readable loan datasets that institutional investors can analyze in real time, improving transparency and capital allocation in fragmented housing markets.
ORCA Computing — a UK-based quantum computing company developing photonic quantum architecture aimed at advancing core hardware capability before large-scale commercial applications emerge. Its focus on photonic systems reflects a belief that scalable hardware design will determine whether quantum computing can move from laboratory demonstrations to enterprise-grade performance.
For Georgiades, the unifying thread across these investments is orchestration — the ability to coordinate data, identity, and permissions across fragmented systems.
Whether in blockchain networks or multi-cloud enterprise stacks, the challenge is similar: information constantly moves between platforms that were never designed to communicate with one another. Without a governing layer that defines who can access what, under which conditions, and with what traceability, autonomy quickly becomes liability.
Enterprises now operate across public clouds, private servers, legacy databases, and third-party software ecosystems. Autonomous systems cannot function safely in that environment unless permissions, data flows, and decision boundaries are engineered into the architecture itself. In that sense, AI investment increasingly resembles enterprise IT modernization — rebuilding the plumbing before turning on the tap.
Identity Orchestration Race
As enterprises accelerate AI deployment, identity orchestration has become a central focus. The rise of autonomous corporate actors is no longer theoretical. Georgiades pointed to emerging experiments in which capital is allocated to AI agents tasked with identifying markets, forming strategies, and launching products with minimal human oversight.
“How do you run a company and deploy it where there’s not an idea right now what the company’s going to do, but there’s a smart autonomous actor who literally does the market research, creates the idea, develops the initial product itself, and then develops a go-to-market strategy? Once it passes a certain number of tests, then you bring in humans to execute part of it,” he said.
“You have an agent that wants to interact with Google and Microsoft. You don’t want it to interact with Amazon for compliance reasons. How do I take my agent that I want to go out in the world, but give it the guard rails and safeguards to ensure that I stay compliant across all these measures?”
Traditional identity and access management systems were designed for human users, while agentic AI introduces non-human actors capable of executing transactions, modifying data, and triggering workflows autonomously. Without real-time oversight and policy enforcement, enterprises risk security breaches, regulatory violations, or unintended operational consequences.
Addressing that gap has given rise to a new category of infrastructure companies focused specifically on identity orchestration, an area where Innovating Capital has invested through its backing of Strata.
Strata Identity — an identity orchestration provider that enables enterprises to bridge legacy and cloud log-in systems without rewriting applications, embedding access control and governance across environments. Its orchestration layer enables interoperability between legacy and cloud systems, allowing large organizations to modernize authentication frameworks without dismantling existing infrastructure, lowering both technical risk and compliance exposure.
Georgiades expects the category to expand as regulators demand clearer audit trails for automated decision-making.
He referenced emerging models such as Feltsense, a startup that has recently raised $5.1 million to fund AI agents designed to run companies autonomously. In that model, investment capital is allocated to software agents that conduct market research, develop products, and design go-to-market strategies before human operators step in.
For Georgiades, such experiments underscore why identity, governance, and orchestration layers must mature alongside increasingly capable autonomous systems.
He said banks, retailers, and logistics providers will increasingly require AI systems that can demonstrate who initiated a transaction, under what permissions, and within which regulatory framework. In highly regulated industries, explainability and traceability are becoming competitive advantages.
Discipline and Convergence
Despite his enthusiasm for AI, Georgiades rejects the notion that crypto has reached maturity.
“Crypto technology isn’t mature. If crypto were mature, we would not be talking about stablecoins. They would be primitives, not products,” he said.
He believes stablecoins, tokenized assets, and blockchain-based data registries must evolve into invisible infrastructure before the sector can claim true maturity. The same logic applies to AI: visibility decreases as reliability increases.
“We develop a particular thesis around a point of technological disruption or differentiation, and then we look at how that thesis can apply to different end markets and sectors, namely, enterprise, SaaS, cybersecurity, Web3, infrastructure, and data infrastructure,” he said. ‘When we go out in the world and look for founders and team operators that are building on that thesis, we give them money and work alongside them.”
That convergence of AI and Web3 infrastructure defines the firm’s forward outlook. Rather than betting on isolated applications, Innovating Capital is positioning itself at the intersection of autonomous systems, standardized data, programmable financial rails, and secure identity layers.
For Georgiades, the next wave of AI investment will not be determined by which chatbot gains the most users, but by which infrastructure providers enable enterprises to trust autonomous systems with critical decisions.
In that environment, capital flows toward control, compliance, interoperability, and orchestration — the architecture that transforms experimental AI into operational reality.