"Most of the industrial world still runs on email, spreadsheets, and laggy screens—but customers now expect Amazon-speed responses," said Kunal Barman, co-founder and CEO of Faction, an AI company building automation agents for manufacturers and distributors. "The gap between what legacy systems can deliver and what the physical economy now demands is exactly where this next generation of automation has to go to work."
In the slow-to-transition world of pipes, valves, fittings, wire, HVAC components, and safety gears, that gap has become more than a matter of convenience. It is increasingly a competitive fault line, separating large distributors that can stitch together their disparate systems with AI-driven agents from those still relying on manual keystrokes and after-hours spreadsheet marathons. As different investments pour into new factories, logistics networks, and infrastructure worldwide, the software layer that orchestrates orders, pricing, and cash flow has emerged as a constraint for profit growth.
A Quiet Revolution in Distribution
Industrial distribution is enormous and structurally fragmented, with the global market projected to reach the tens of trillions of dollars by 2030, growing steadily as capital flows into manufacturing, energy, and construction. Yet, inside many of these businesses, routine commercial work—such as quoting, order entry, sourcing, price checks, and collections calls—still relies on humans retyping data from PDFs and emails into CRM systems that are not well-suited for today's speed and complexity.
Earlier waves of automation focused on plant-floor robotics or narrow workflow scripts rather than the full quote‑to‑cash lifecycle that distributors live and die by. Over the past two years, however, AI agents that can read documents, interact with customers via voice, and act directly within enterprise software have transitioned from experimentation to deployment, particularly in sectors such as PVF, electrical, and HVAC, where product catalogs are vast, and orders are complex. For a growing cohort of distributors, these agents are becoming the connection between legacy infrastructure and modern expectations.
Barman describes Faction's role in that shift in simple terms: "We build agents that sit on top of ERP and CRM, watch the work humans are doing, then start doing 20%, then 40%, then 60% of it without asking the customer to rip anything out."
The New Automation Stack
What distinguishes this "next generation" of automation from the last is not only that AI models have become far better at language and pattern recognition, but that they can work across different modes: documents, structured data, and voice. A quote request may arrive as a messy spreadsheet attachment, a half-complete email, or a phone call from a jobsite; the same agentic infrastructure routes, interprets, and executes the workflow end‑to‑end. Faction, for instance, markets a multimodal platform that combines document understanding with voice agents and proprietary data models tuned to distributors' specific product universes.
Across early adopters, Barman discusses, the results cluster around a few recurring numbers: a 75% reduction in manual data entry, automation of 15 to 25% of sales and support workload, and increases in revenue per sales representative as staff spend more time on selling rather than data cleanup. Those figures align with broader industry research, which finds that AI-powered quote-to-cash systems can significantly reduce average processing time and minimize revenue leakage. In sectors with complex pricing and high transaction volumes, the cumulative impact on profitability can be substantial.
Behind these agents is a more layered stack than the customer sees. Underneath the conversational interfaces sit integration connectors into ERP and CRM platforms, data enrichment pipelines that normalize messy product information, and policy engines that encode rules for pricing, credit, and approvals. Around that, observability tools monitor which tasks the agents are handling, where they defer to humans, and how often exceptions occur. For distributors accustomed to flying blind on workflow performance, this instrumentation alone can be a revelation.
A Market Tilting Toward AI Agents
Industrial automation more broadly is on a parallel trajectory, with global industrial automation and control systems already worth hundreds of billions of dollars and projected to grow strongly through 2030. While much of that spending still goes to hardware and control software on the factory floor, a growing share is being allocated to systems that coordinate orders, inventories, and financial flows across the supply chain.
Financial leaders have taken notice. Surveys indicate that a large majority of companies are investing in AI to improve cash flow and working capital, often starting with invoice processing, collections, and revenue recognition. In that context, a distributor's decision to deploy agents to handle collection calls or reconcile disputed invoices appears more like a pragmatic catch-up with the corporate world.
Human Work, Redefined—Not Replaced
Barman stresses that this wave of automation is not, in his perspective, a blunt instrument for labor substitution. "If you walk a warehouse or sit with a branch sales team, you see people working flat out," he said. "The constraint isn't effort; it's that they're spending too much effort on low‑value clicks. The agents are there so those teams can spend more time on growth initiatives and customer relationships, not so they can be replaced."
The logic aligns with broader research on AI-augmented work, which suggests that productivity gains are highest when systems automate repetitive tasks. Still, humans retain control over exceptions, complex negotiations, and relationship-building. For a distributor's inside sales representative, that might mean an AI system that pre-builds quotes, flags cross-sell opportunities, and recommends price bands. In contrast, the representative focuses on understanding project timelines, coordinating with field sales, and solving problems that do not fit the template. When manual data entry shrinks, the job changes shape rather than vanishing.
Skeptics Warn of Overreach
Not everyone is convinced that AI agents are an unalloyed good for the distribution sector. "There's a real risk that in the rush to automate, companies will embed their existing bad processes in code, and then it becomes much harder to change," said Maria Hernandez, a supply‑chain technology consultant who has advised mid‑market distributors on ERP rollouts. "If your pricing logic is inconsistent, your product data is a mess, or your credit policies are ad hoc, an agent can amplify that chaos at machine speed."
Hernandez also worries about over‑reliance on vendors whose platforms may be opaque. "These are probabilistic systems; they will make mistakes, and you need robust guardrails and auditability," she said. "Most distributors do not have the in‑house AI expertise to evaluate or monitor vendor models properly, and that asymmetry could become a problem if something goes wrong."
Founders Shaped by the Physical Economy
Barman's conviction that the physical economy needs smarter software predates the generative AI surge. At BlackRock's Financial Markets Advisory unit, he worked on the design of Saudi Arabia's USD 53 billion National Infrastructure Fund, helping to define investment strategies across energy, logistics, and industrial development—an immersion in large-scale, capital-intensive systems where seemingly small operational frictions can have outsized macroeconomic consequences. Earlier roles in investment banking and venture capital exposed him to both deal‑making at scale and the emerging generation of AI‑native startups.
Those experiences inform his view that industrial distributors, often overlooked in technology narratives, occupy a pivotal point in the global economy. "You can build all the new factories and warehouses you like," he said, "but if the distributors feeding them are stuck on 20‑year‑old systems, the throughput of the whole system is constrained." In his telling, the right kind of automation is less about polishing back‑office efficiency metrics and more about allowing the physical world to move at the speed that new capital investment makes possible.
Faction's focus on U.S. enterprise distributors in sectors such as PVF, electrical, HVAC, packaging, and safety reflects that thesis. These are categories where product variance is high, downtime is costly, and customer relationships remain heavily relationship-driven. If the agents do their job quietly, the most visible changes may not be in flashy dashboards, but rather in fewer errors, faster replies, and account managers with more time to think.
Profitable Growth in a Constrained World
Behind the enthusiasm for AI agents lies a harsher macro reality: labor markets in many industrial regions are tight, capital is not freely available, and customers are demanding both lower prices and more flexible service. Under those conditions, profitable growth depends on extracting more value from existing assets rather than simply adding more of each. Automation that can compress cycle times, reduce clerical workload, and improve decision quality offers one of the few levers left that does not require massive new capital spending.
Forecasts anticipate that by late this decade, a significant share of business decisions in large enterprises will be at least partially automated or augmented by AI agents. In that world, distributors that have not begun the journey risk finding themselves structurally less responsive than their peers. Conversely, those that move early but without discipline may discover that poorly governed agents can introduce new forms of operational risk even as they solve old ones.
Barman frames the trade‑offs in pragmatic rather than utopian terms. "The distribution industry has always been about thin margins and tight relationships," he said. "What changes now is that the operational baseline for what 'good' looks like is rising fast. If you can answer complex RFQs in hours instead of days, give accurate availability across a messy network, and get paid faster without hounding customers, that shows up directly in the P&L." For him, the question is less whether automation will reshape the sector than how evenly its benefits will be distributed.
When asked how this story might read in 2030, Barman offered a characteristically measured answer. "If we get this right," he said, "no one will talk about AI agents as a separate thing. They'll just talk about distributors that are easier to do business with, that never seem to drop the ball, and that keep investing in their people even as they scale."
"The real test is whether the people on the warehouse floor, in purchasing teams and at branch counters feel like their work got better—more creative, more relationship‑driven—because the drudgery went away," Barman stated. In the long shadow of new factories, ports, and power lines, it is that quieter recalibration of everyday work, mediated by lines of code that few will ever see, that may determine whether the next generation of automation truly becomes a game changer for enterprise distributors, and for the physical economy they help keep moving.
ⓒ 2025 TECHTIMES.com All rights reserved. Do not reproduce without permission.
