New AI 'Maven' Aims to Unlock Decades of Buried Materials Knowledge

TEL AVIV, Israel – April 29, 2026 – MaterialsZone, a materials informatics company, today launched Maven, a conversational AI interface designed to tackle one of the most persistent challenges in industrial research and development: accessing and making sense of decades of accumulated, often siloed, corporate knowledge. The new agentic AI capability, integrated into the company's enterprise platform, aims to provide a faster, more intuitive way for R&D and production teams to search, visualize, and analyze vast datasets through simple, natural language.

By combining generative AI's conversational abilities with deep analytical AI, Maven seeks to empower organizations to finally leverage their own proprietary data, a hurdle that has caused a staggering number of enterprise AI initiatives to fail.

The 95% Problem: AI's Struggle in Heavy Industry

Despite billions invested in enterprise AI, the path to a positive return on investment has been fraught with difficulty. A recent, sobering MIT study highlighted this reality, finding that approximately 95% of enterprise AI pilots fail to deliver measurable impact. The report identified the core barrier not as a flaw in the AI models themselves, but in the inability to securely connect them to proprietary systems and learn from the specific context of an organization's daily operations.

The study painted an especially bleak picture for heavy industry, describing the Energy and Materials sectors as having "near-zero adoption; minimal experimentation." This lag is often attributed to the complex, highly specialized nature of the data and the immense security concerns surrounding valuable intellectual property. Generic AI tools, while useful for general tasks, frequently fail when confronted with the intricate workflows and unique data structures of materials science.

MaterialsZone aims to confront this challenge directly with Maven. The platform is engineered to bridge the gap identified by MIT, providing a secure environment where the AI can be deeply integrated with a company's accumulated knowledge—from experimental results and formulation data to supplier specifications and production metrics.

A Digital Expert for Every Scientist

At the heart of Maven's design is the ambition to digitize and scale institutional memory. Often, the most valuable knowledge in a large R&D organization resides in the mind of a single, long-tenured expert who remembers decades of experiments. Maven is designed to be that expert, but accessible to everyone on the team.

"Across the organizations we work with, there's often the one expert, the person people naturally come to for advice and insight, who's been around for decades and remembers every experiment, every formulation, every result," said Ori Yudilevich, CPO of MaterialsZone, in the company's announcement. "Maven captures that expertise and makes it accessible to the entire team through a simple conversation."

Technically, Maven functions as a large language model (LLM) layer built upon MaterialsZone's existing materials informatics platform, which already specializes in data management and machine learning for R&D. This integration allows users to interact with their complex data without writing a single line of code. A researcher can now simply ask questions like, "Show me all formulations from the last five years that used polymer X and achieved a tensile strength above Y," or "Visualize the performance of our current product compared to competitor Z's technical data sheet." Maven can then instantly analyze the data, generate charts, and surface the relevant past experiments, effectively turning hours or days of data-hunting into a minutes-long conversation.

Extending Intelligence Beyond the Lab

The impact of such a tool extends far beyond saving time for individual researchers. The company reports that scientists can spend up to 60 percent of their time on knowledge retrieval tasks—fielding internal questions, locating past results, and verifying data. By automating these queries, Maven promises to free up highly skilled personnel to focus on actual discovery and innovation.

Furthermore, the platform is designed to break down inter-departmental silos. Sales teams, for instance, can use Maven to instantly check if their company's materials can meet a potential customer's specific requirements, generating a data-backed answer without having to wait days for an R&D response. Similarly, Procurement teams can evaluate potential raw material substitutions by querying Maven about their impact on existing formulation requirements, leading to faster, more informed, and potentially cost-saving business decisions.

Securing the Crown Jewels

For any enterprise dealing with proprietary formulations and trade secrets, the primary concern with adopting a new AI platform is data security. MaterialsZone is addressing this head-on by building its security framework around data isolation. The company, which is SOC 2 Type II compliant, ensures that each customer's data is stored exclusively within that customer's own secure environment.

Crucially, this data is used only to train that specific customer's proprietary AI models. The platform's architecture prevents data from one organization from ever being used to train shared models, meaning no company can access or inadvertently benefit from another's intellectual property. A granular, role-based permission system adds another layer of control, ensuring that individual users within an organization can only see the data and analyses they are cleared to access.

A Glimpse into the Future of R&D

The launch of Maven comes at a time when the materials industry is actively seeking ways to accelerate innovation. The field is seeing a rise in specialized AI platforms, with companies like Citrine Informatics also offering AI-driven solutions to speed up product development. The common goal is to shift materials R&D from a slow, trial-and-error process to a more predictive, data-centric cycle.

By integrating a powerful conversational interface with a secure, specialized data platform, MaterialsZone is betting that the key to unlocking AI's potential lies in making it both accessible and trustworthy. The company will be demonstrating Maven's capabilities to scientists and industry professionals at several upcoming trade shows, including The American Coatings Show in Indianapolis and Chemspec Europe in Cologne, Germany, offering a first-hand look at what could be the future of materials innovation.