Aerium's AI Agents Are Negotiating With Suppliers for Industrial Manufacturers

If you want to sell software to an industrial manufacturer, you better not start by telling them to rip out their existing systems. Aerium, a Toronto-based startup, is taking the opposite approach, building what it calls an AI workforce that plugs into the data and tools already running a factory or distribution center. The company’s bet is that the path to automating complex supply chain, operations, and commercial tasks lies not in a new hardware deployment, but in connecting, cleaning, and centralizing the operational data those businesses already generate [BetaKit, October 2026].

In July 2026, Aerium secured a $2.5 million USD seed round, led by New York’s Gutter Capital, to develop that platform further [BetaKit, October 2026]. The financing, structured as a simple agreement for future equity, suggests investors are buying into the premise that industrial AI can be deployed as a software layer, not a physical overhaul. For a sector known for long sales cycles and capital-intensive upgrades, that is a pragmatic wedge.

The bet on existing data

The core of Aerium’s proposition is its insistence on using a manufacturer’s incumbent systems. The platform is designed to connect to a customer’s ERP, MES, and other operational software, gather and clean the data flowing through them, and centralize it into a single source [BetaKit, October 2026]. On top of that unified data layer, the company deploys AI agents and machine-learning models to monitor operations and automate specific tasks.

The scope of those tasks is ambitious, spanning four traditional business functions. On its website, Aerium lists capabilities across supply chain, operations, commercial, and back-office activities [Aerium, retrieved 2024]. The most concrete examples involve AI agents that monitor market prices for thousands of SKUs, identify savings opportunities, and can even autonomously negotiate with suppliers [joinaerium.com, retrieved 2026]. Other promised features include dynamic job scheduling, root-cause analysis for downtime, and automated quoting and margin analysis.

The team and the check

The founding team brings a blend of consulting pedigree and claimed sector familiarity. CEO Adam Trotman is a McKinsey alum, a background that typically signals an understanding of large-organization processes and buying committees [LinkedIn, retrieved 2026]. Co-founder Michael Dawes’s public profile is less detailed, though a Bloomberg listing identifies him as a senior vice president at Equitable Life Insurance Company of Canada [Bloomberg Markets, retrieved 2026]. In a statement to BetaKit, Trotman said both founders were exposed to manufacturing from a young age due to their parents working in the sector [BetaKit, October 2026].

The $2.5 million seed provides runway to turn the platform vision into a saleable product. The round was led by Gutter Capital, with participation from undisclosed strategic angel investors [BetaKit, October 2026]. The company has also been part of the Entrepreneurs Roundtable Accelerator program in New York [Entrepreneurs Roundtable Accelerator, retrieved 2026].

Seed Round (July 2026) | 2.5 | M USD

The realistic competitive set

Aerium’s stated ambition places it in a crowded but fragmented field. It is not competing with monolithic ERP vendors like SAP or Oracle on core system-of-record functionality. Instead, its realistic rivals are the point solutions and emerging AI platforms targeting specific operational pain points within industrial settings. The competitive landscape breaks down into a few clear categories.

• Vertical SaaS incumbents. Companies like ProShop (for discrete manufacturing) or Fishbowl (for inventory) own the workflow layer for many small to mid-sized manufacturers. Aerium would need to integrate with them, not replace them.
• Supply chain intelligence platforms. Tools like Keelvar or Jaggaer focus on sourcing and procurement automation, a surface Aerium explicitly targets with its autonomous negotiation claims.
• Industrial IoT and analytics. Players like Falkonry or Sight Machine use sensor data for predictive maintenance and production optimization, overlapping with Aerium’s operations monitoring pitch.
• Generic AI automation. Broad-based robotic process automation (RPA) and AI agent platforms from UiPath or Microsoft could be configured for similar back-office tasks, though they lack industry-specific data models.

Aerium’s differentiation rests on bundling these capabilities,supply chain, operations, commercial,into a single, industry-tailored platform that promises to improve the P&L across multiple functions [Farm Equipment Manufacturers Association, retrieved 2026]. Its success hinges on proving that this integrated approach delivers more value than stitching together best-of-breed point solutions.

Where the wheels could come off

For all its pragmatic positioning, Aerium faces the classic execution challenges of any early-stage enterprise AI company. The platform’s value is contingent on deep, clean integrations with a wide array of legacy industrial systems, which are notorious for their complexity and data silos. The company has not publicly named any design partners or pilot customers, so the real-world friction of these integrations remains unproven [BetaKit, October 2026].

Furthermore, the concept of AI agents autonomously negotiating with suppliers is a bold claim that sits at the frontier of what procurement teams are comfortable delegating. Trust, liability, and the need for human-in-the-loop oversight could significantly slow adoption of that specific feature, regardless of its technical feasibility. The sales motion for a platform this broad may also be complicated, as it potentially touches budget owners in procurement, operations, and sales departments within a single customer organization.

The next twelve months

The immediate roadmap is clear: deploy the capital to advance the product and secure initial customers. The ideal customer profile is likely a mid-market industrial manufacturer or distributor with existing digital systems, complex supply chains, and enough operational data to feed the AI models. These are companies large enough to have pain points across multiple functions but agile enough to pilot new software without a two-year procurement cycle.

Aerium’s membership in the Farm Equipment Manufacturers Association suggests an early focus on that vertical, which aligns with the founders’ stated background [Farm Equipment Manufacturers Association, retrieved 2026]. A credible milestone for the next year would be announcing a handful of paid deployments, particularly in specialty chemicals or metal fabrication, two other sub-verticals mentioned on its site [Aerium, retrieved 2024]. Concrete case studies showing quantifiable improvements in metrics like on-time delivery, material cost, or quote turnaround time would be the strongest possible validation that its AI workforce can deliver on the P&L promise.

Sources

1. [BetaKit, October 2026] McKinsey alumni raise $2.5 million USD to build AI for industrial manufacturers | https://betakit.com/mckinsey-alumni-raise-2-5-million-usd-to-build-ai-for-industrial-manufacturers/
2. [Aerium, retrieved 2024] Aerium | The AI workforce for industrials | https://www.aeriumhq.com/
3. [LinkedIn, retrieved 2026] Adam Trotman - Co-founder @ Aerium | McKinsey alum | https://ca.linkedin.com/in/adamtrotman
4. [Bloomberg Markets, retrieved 2026] Michael Dawe, Equitable Life Ins Co of Can: Profile and Biography | https://www.bloomberg.com/profile/person/17662101
5. [joinaerium.com, retrieved 2026] Aerium | Procurement intelligence | https://www.joinaerium.com/
6. [Farm Equipment Manufacturers Association, retrieved 2026] New Member Profile: Aerium | https://www.farmequip.org/news/new-member-profile-aerium
7. [Entrepreneurs Roundtable Accelerator, retrieved 2026] Companies - Entrepreneurs Roundtable Accelerator | https://www.eranyc.com/companies/