Metalshub AI: Commercial Intelligence through Data Governance

Metalshub is bringing structured commercial context to enterprise AI in raw materials, transforming daily transactional data into a secure, governance-first asset for procurement and sales leaders.

Across raw materials markets, digital decision-makers and commercial leaders are evaluating how to apply artificial intelligence within their sales, trading and procurement operations. Much of the broader industry conversation centres on selecting the right large language model (LLM). Model performance, however, is not the main bottleneck. The true driver of value is access to structured, domain-specific business context.

Metalshub’s AI vision addresses this fundamental requirement directly. Structured commercial data provides the context layer that enables AI to support complex raw materials decisions. By focusing on data governance and two-sided verification, Metalshub turns everyday commercial interactions into an institutional asset that organisations can safely expose to AI tools.

In physical raw materials trading, commercial teams generate deep operational context every day through technical specifications, price enquiries, and multi-round negotiations. Enterprise resource planning (ERP) systems excel at recording transactional results, such as purchase orders and delivery confirmations. However, when no one structures the intermediate communication, the commercial intent behind those outcomes gets discarded.

The reasoning behind price adjustments, technical specification trade-offs, and counter-offer histories rarely make it into formal system records. That context often lives in someone’s head, or nowhere at all. When experienced category managers or sales managers leave an organisation, years of institutional memory exit with them, taking critical historical perspective on counterparty flexibility, quality tolerances, and negotiation leverage out the door.

Without this structured history captured in a central system, any incoming AI tool is forced to work with incomplete facts, making it impossible to reconstruct why decisions were made or reproduce successful strategies.

To capture this intent without adding manual workload, Metalshub AI provides Text2Platform (i.e., text-to-platform AI functionality). For example, a commercial manager can forward an email quote or push a quick WhatsApp message directly to Metalshub to capture an enquiry and submit an offer. This converts conversational detail into structured platform records from the start.

Metalshub captures operational context under the framework of Commercial Intelligence. By managing transactions across an active two-sided network, the platform captures specification details, enquiry patterns, and negotiation steps as standard workflow.

Network data remains trustworthy because both buyers and sellers confirm the exact same terms at the moment of creation. Rather than attempting to reconstruct deal history from unverified, free-form email threads after a transaction closes, a live network records counter-offers, terms, and specifications as verified data. Metalshub acts as the system of record for this Commercial Intelligence, whether accessed natively within the platform or utilised inside a customer’s proprietary AI environment.

To meet varying technical capabilities and corporate requirements, implementation follows a dual-track strategy:

• Enterprise Integration via MCP: Enterprise teams with established internal AI initiatives can connect Metalshub context directly to their own LLMs and custom digital assistants using the MCP. By combining Metalshub’s real-time commercial data with internal CRM, ERP, and market intelligence feeds, AI agents gain full contextual awareness. For instance, connected agents can run complex queries across datasets to analyse operational patterns, such as identifying which raw material suffers from the weakest average offer rate per enquiry or the lowest supplier response rate. Such integration enables them to perform complex analysis, generate market insights, and execute instructed tasks within the customer’s secure environment.
• Native Platform Capabilities: For teams seeking ready-to-use capabilities without building custom AI infrastructure, native co-pilots and agents operate directly within the Metalshub interface. These embedded capabilities assist users in real time during daily tasks, helping draft RFQs, analyse historical pricing during active negotiations, and identify suitable counterparties.

Strict data governance and complete customer control remain central to this framework. Customers retain absolute ownership over their proprietary commercial data, determining exactly which datasets can be accessed, how they are routed, and which internal or external systems are permitted to interact with them. Metalshub enforces strict enterprise-grade permissions to ensure sensitive negotiation histories and commercial terms are never used to train public models or exposed across organisational boundaries.

Furthermore, AI operations focus on tasks where the value generated justifies the computing expense. For high-risk operational steps or monetary transactions where errors carry severe risk, deterministic logic and human oversight remain mandatory. For instance, AI agents can analyse past negotiation terms to draft Round 1 counter-offers based on ideal historical patterns. To eliminate the risk of hallucinated actions or unauthorised commitments, these workflows require a human user to review and approve the draft before any offer is sent.

Applied effectively, commercial intelligence generates clear practical outcomes:

• Better counterparty recommendations derived from recorded performance history
• Negotiation strategies grounded in verified historical contract terms
• Preserved institutional memory that survives team transitions

In raw materials markets, long-term advantage with AI comes down to structured context and rigorous data governance rather than model selection. By maintaining strict data privacy while creating secure connections into existing enterprise IT stacks, sales and procurement teams build a permanent operational foundation capable of supporting autonomous workflows and intelligent decision-making for years to come.

Our platform roadmap follows a clear progression. It began with our AI Text-to-Platform capabilities, which automatically convert unstructured email quotes into structured data inside the system. Building on this core platform structure, we are deploying live MCP connections, allowing internal enterprise AI environments to query and retrieve this verified context directly. Development continues toward specialised agent workflows, including negotiation assistance, with several sophisticated customers already conducting AI pilot projects using these capabilities.

We remain focused on helping industry leaders translate raw commercial activity into measurable bottom-line value. To explore this vision further and see how leading metals and mining enterprises are preparing their data stacks for the AI era, join our upcoming MetalshubTalks episode on “The State of AI for Buying and Selling Raw Materials.” In this session, our leadership team will discuss how to bridge the gap between transactional systems and enterprise AI platforms, demonstrating practical use cases and setting the stage for future AI-driven commercial operations.