Avent Is Automating the Part of Industrial Sales That Nobody Wants to Do

The Macro: Industrial Distribution Is a Trillion-Dollar Industry Running on Fax Machines

I am going to describe a workflow and you tell me what decade it is from. A customer sends a request for quote via fax. A sales rep reads the fax, opens an ERP system, manually searches the product catalog for each line item, checks pricing against historical orders, types the quote into the system, and sends it back via email. The customer sends a purchase order. A different person reads the PO, re-enters every line item into the ERP, validates quantities and pricing, and creates a sales order.

This is happening right now, in 2026, at thousands of industrial distributors across North America. It is not a small market. Industrial distribution in the US alone is roughly $7 trillion. The companies that move pipe fittings, electrical components, bearings, fasteners, and industrial chemicals are large businesses with thin margins and high transaction volumes. And they are running their sales operations on processes that have not fundamentally changed since the 1980s.

The reason is not stupidity. These are smart operators who have survived decades of competition. The reason is that industrial catalogs are enormous (hundreds of thousands of SKUs), pricing is complex (customer-specific, volume-tiered, contract-dependent), and the existing ERP systems were not designed for automation. Epicor, Infor, SAP. These platforms were built as databases with forms on top. They expect a human to sit at a keyboard.

Previous attempts to modernize industrial distribution have mostly focused on e-commerce. Faire, for wholesale. Various B2B marketplace plays. But the actual bottleneck is not how orders arrive. It is what happens after they arrive. The quoting and order entry process is where the labor cost and error rate live. A sales rep at a mid-size distributor might process 30 to 50 quotes per day. Each one takes 10 to 20 minutes. That is an entire person’s day spent doing data entry instead of building customer relationships.

The Micro: A Berkeley Engineer Who Worked on the Factory Floor

Avent builds AI sales reps for industrial distributors and manufacturers. The system processes customer requests across email, fax, and phone. It reads RFQs, generates quotes using historical pricing and product knowledge, creates sales orders in the ERP, and even handles procurement by sourcing unavailable items from partner distributors. The company says it is already generating over $200,000 worth of quotes every day for distributors across North America and the UK.

Abhay Kalra is the founder and CEO. He studied Electrical Engineering and Computer Science at UC Berkeley, where he did computer vision research for manufacturing defect detection. More importantly, he actually worked as an operator in manufacturing and industrial distribution. That hands-on experience matters in a market where the customers are deeply skeptical of technology vendors who have never set foot in a warehouse.

The technical architecture is layered. A data layer aggregates ERP, CRM, and catalog information in real time. An integration layer connects to the existing systems via APIs. A knowledge layer captures business rules, pricing logic, and customer-specific context. An action layer generates quotes and orders. An intelligence layer processes documents with AI reasoning. The integrations list includes Epicor, Oracle NetSuite, SAP, Infor, Microsoft Dynamics 365, and Salesforce.

That integration list is critical. In industrial distribution, the ERP is the center of the universe. Nothing matters if you cannot read from and write to the ERP. The fact that Avent supports the major platforms means they can sell to a broad swath of the market without custom integration work for every deployment.

It is a two-person team. They went through Y Combinator’s Summer 2025 batch. The product has a demo and login portal, suggesting active customers beyond the pilot stage. The $200,000 in daily quotes is a strong number for a company this young. That implies real order volume flowing through the system, not just a proof of concept.

The competitive landscape includes Proton.ai, which does AI-powered recommendations for distributors, and various RPA tools that automate specific ERP workflows. But nobody else is doing end-to-end quote-to-order automation specifically for industrial commerce. The niche is narrow and deep, which is exactly the kind of market where a focused startup can build a real business before the incumbents notice.

The Verdict

I think Avent is one of the more interesting vertical AI plays I have seen this year. Industrial distribution is massive, underserved by technology, and full of repetitive processes that are perfect for automation. The $200,000 in daily quote volume suggests they have already found product-market fit with early customers, which puts them ahead of most companies at this stage.

The risk is the long sales cycle. Industrial distributors are conservative buyers. They do not adopt new software quickly, especially software that touches their ERP and pricing data. Building trust with these customers takes time, and a two-person team has limited capacity for the on-site visits and hand-holding that enterprise industrial sales typically require.

In thirty days, I want to see how many distributors are running Avent in production and what their error rate looks like. A bad quote in industrial distribution is not just a lost deal. It is a customer calling to yell at you. In sixty days, I want to know whether the fax and phone processing is reliable or whether most real usage comes through email. In ninety days, the question is whether Avent can scale to large distributors with complex multi-location operations, or whether the product works best for mid-market companies with simpler setups. The market is there. The product is there. The execution challenge is convincing an industry that still uses fax machines to trust an AI with their quoting.