By David Arnež, co-founder at Inflectiv.ai

In Crypto Theses 2026, Messari makes a simple but powerful observation: the most important AI projects in crypto won’t be the ones selling raw compute or generic inference. Instead, they will be the ones that create, structure, and distribute scarce, high-value data.

→ Learn more

As public data dries up and AI development shifts toward post-training, reinforcement learning, and physical AI (like robotics), data becomes the real bottleneck. Messari calls the systems that solve this problem decentralized data foundries — protocols that crowdsource, curate, and monetize specialized datasets.

Inflectiv fits directly into this picture. Not as a theoretical match, but as a practical implementation of what a modern data foundry looks like, one that goes beyond raw data collection and into structured intelligence.

From Raw Data to Structured Intelligence

One of Messari’s key insights is that more data isn’t the answer. What matters is:

• Data quality,

• clear structure,

• fast turnaround, and

• immediate usability.

AI labs are already flooded with low-signal datasets. What they actually want are datasets that are:

• Domain-specific,

• queryable,

• Provenance-aware, and

• ready to plug into training or inference pipelines.

This is where Inflectiv differs from most data platforms. Inflectiv doesn’t stop at collecting data. It takes messy inputs → documents, SOPs, research archives, sensor streams, trading signals and turns them into structured, agent-ready intelligence.

You can think of Inflectiv as not just a ‘data foundry’:

• Raw inputs go in, structured, and standardized intelligence comes out.

That structure is what makes data usable, reusable, and valuable. Rather than focusing on one narrow data type, Inflectiv acts as a universal structuring layer. Any high-value domain; robotics, finance, compliance, research can be transformed into intelligence that machines can actually use.

Robotics, DePAI, and the physical AI bottleneck

Messari’s strongest data argument appears in robotics and embodied AI. Physical AI systems need millions of hours of real-world data to improve. Today, most robotics datasets sit in the low thousands of hours - orders of magnitude short of what’s needed.

Because of this gap, Messari highlights DePAI (Decentralized Physical AI) networks as one of the most promising areas in crypto AI. These systems focus on collecting real-world data such as:

• Navigation data,

• motion and telematics,

• audio and video, and

• environmental and sensor signals.

Inflectiv aligns closely with this need. Through its sensor-native data model and the SenseNet approach, Inflectiv will allow globally distributed devices to contribute structured perception data, not just raw logs.

That distinction matters. Robotics teams don’t want unfiltered sensor exhaust. They need datasets that are:

• Linked to specific tasks,

• auditable and verifiable, and

• clearly licensed and reusable.

• 
  

Inflectiv’s dataset model embeds consent, attribution, and provenance directly into the data itself - exactly the qualities Messari identifies as necessary for serious AI and enterprise adoption.

While many DePAI projects focus purely on collection, Inflectiv goes further by making that data liquid, composable, and reusable across agents and workflows.

The data economy as a flywheel, not one-off sales

Another important idea in the Messari report is that the best data platforms don’t rely on one-time sales. Trust is built slowly (step by step OR in Inlfectiv’s case; dataset by dataset) but value compounds when data is:

• Reused,

• queried repeatedly, and

• embedded into live systems.

Inflectiv is designed around this reality. Instead of selling static datasets, it enables a complete economy:

• Dataset monetization through tokenization or per-query usage,

• agent-level integrations, and

• ongoing downstream applications.

This turns data into something closer to living infrastructure rather than a static product.

In this model, Inflectiv doesn’t compete with AI labs. It becomes part of their supply chain, providing the incentives and tooling that let contributors continuously improve datasets while ensuring buyers only pay for real usage.

Beyond compute: why Inflectiv fits the 2026 thesis

Messari is clear that decentralized compute alone hasn’t delivered lasting value. Compute tends to become commoditized. The long-term winners are those that control scarce inputs. Today, that scarce input is structured, domain-specific intelligence.

Inflectiv fits this future because it:

• Treats data as something that can be refined, not just collected,

• prioritizes structure, context, and usability over raw volume,

• aligns incentives between data contributors and data users, and

• focuses on frontier needs like robotics, agents, and real-world automation.

In Messari’s framework, decentralized data foundries sit at the center of crypto-AI’s next phase. Inflectiv can be understood as a generalized intelligence foundry (a system that turns experience, expertise, and perception into usable intelligence markets).

If compute defined the previous cycle, structured intelligence defines the next. Inflectiv is already being built for that world.

David is a co-founder of Inflectiv and brings over a decade of experience across Web2 and Web3 companies. His background spans the full company-building lifecycle, from early idea formation through execution and scale. He has completed two full startup cycles in Web2, with a focus on IoT and SaaS businesses.

Alongside his industry work, David is an assistant professor and a PhD candidate, where his research focuses on entrepreneurial behavior, artificial intelligence, data systems, and applied research. His academic work informs his practical approach to technology adoption and market design.

At Inflectiv, David serves as CBDO, leading partnerships and ecosystem developments. His work centers on the real-world deployment of structured data for AI agents and automation, bridging technical capabilities with concrete business use cases.

Learn more about Inflectiv: inflectiv.ai

Follow us on X: Official Inflectiv X Account