Last month Lantern, an AI-powered, inventory forecasting system, joined NAED. The association announced their membership a couple of weeks ago.

Having met Conor Leen almost two years ago, I reached out to him for an update and to offer him the opportunity to share more with the electrical industry.

The company has had good success in the HVAC industry and one of Channel Marketing Group’s electrical distribution clients works with the company.

While there isn’t an NAED meeting until mid-January to meet them, I would encourage distributors to reach out to learn more. They have a different approach and are helping distributors better manage their largest non-people asset … their inventory.

What makes Lantern a little different is that leadership has a passion for distribution. Further, while they are a start-up, they’ve now also raised money twice with the second time being from a couple of VC firms … so they’ve gone through some vetting.

Lantern Brings Machine Learning Forecasting to Electrical Distribution

Who is Lantern

Lantern was founded by Matt Rojas, Conor Leen, and Jay Shitole, Stanford graduates whose team spans alumni of Google, Amazon, Stripe, and Walmart, and whose ties to distribution run deeper than their resumes. Rojas’s father worked in distribution, so the industry has been personal from the start. At Stanford, the founders wrote their thesis on distribution, interviewing more than 200 industry executives and ending every conversation with the same question: what is the biggest challenge you are facing right now? One answer came up every single time. Forecasting.

That research became Lantern, an AI forecasting and purchasing platform for wholesale distributors. The company launched publicly in November 2025 and recently announced a $7 million seed round led by Primary Venture Partners and Pear VC to hire engineers, accelerate product development, and expand the machine learning behind its recommendation engine.

Why Lantern Exists

For electrical distributors, inventory is the business. It is the largest asset on the balance sheet, the product customers depend on, and the strongest predictor of success. Yet buyers face an impossible balancing act every week, across thousands of SKUs. Buy too little and you get stockouts, missed sales, and lost customers. Buy too much and you get a mountain of dead inventory in the warehouse. Most distributors end up living with both.

Lantern’s view is that this is not a people problem. Buyers are working harder than ever. The problem is that demand is more volatile than ever, lead times shift constantly, and assortments keep expanding, while purchasing decisions are still guided by rules, averages, and gut feel inside systems that assume the future will look like the past.

What and How

Lantern uses machine learning to forecast demand and lead times, optimize order frequency, rank products, and assign safety stock to hit service level targets, then turns all of it into concrete purchase and transfer recommendations for buyers. The models detect true seasonal patterns, use data across the entire catalog to forecast long tail items with limited history, and distinguish one-time anomalies, such as a large project order, from the start of a sustained demand shift. The platform sits on top of a distributor’s existing ERP, and implementations typically go live in weeks.

Built for How Electrical Actually Works

Generic forecasting tools stumble on the realities of the electrical counter. Lantern built for them directly through a design partnership with Schaedler Yesco:

• Units of measure as products are actually bought and sold. Electrical products move in feet, reels, pieces, and packs, and the buying unit often differs from the selling unit. Lantern handles these conversions natively, referencing the purchasing unit rather than defaulting everything to each, and supports per hundred and per thousand pricing (hardware priced per 100, wire per 1,000) without rounding errors creeping into POs.
• Kits and assemblies. When an item is really a kit of components, Lantern breaks it down to the bill of materials, so demand and inventory are calculated at the component level instead of treating the kit as one opaque SKU.
• Buy pack aware rounding. A recommendation of 70 units for a product that ships in packs of 100 is useless. Lantern rounds recommendations to real buy pack quantities, including smart rules like rounding up to a full pack when the need crosses half the pack size.
• Vendor targets in the units suppliers actually set. Freight minimums and minimum order values are not always dollar based. Lantern tracks progress toward vendor targets in dollars, pounds, or pieces, right on the purchasing screen, so buyers can hit freight allowances and order minimums as they build the PO.
• Flexible buy line and supplier relationships. The rigid model of one buy line, one vendor, one PO does not match electrical purchasing. In Lantern, items within a single buy line can be sourced from multiple suppliers, with automatic supplier defaulting from purchase history and one recommendation splitting into separate POs per vendor, which is essential for miscellaneous supply lines with many manufacturers. Multiple buy lines can be combined onto a single PO when they ship from the same supplier, and whole PO supplier changes take one click instead of line-by-line edits.
• Wire as it physically exists. 2,000 feet on hand might be one master reel or forty short remnants. One can serve a contractor’s continuous run and the other cannot. Lantern distinguishes usable reels from remnant pieces, showing piece counts and individual lengths, so buyers reorder based on what is actually sellable rather than a footage total that hides a stockout.

Expected Outcome

Across live customers, Lantern reports inventory reductions of 20 to 30%, roughly one additional inventory turn, and fill rate improvements of 5 to 10 points, plus hours of manual work returned to buyers each week. R&E Supply, an eight-branch HVAC distributor that also sells some electrical material, cut inventory 34% for about $520,000 in annual savings.

Aside from being a member of NAED, Lantern is a member of other distributor associations.

For electrical distributors who want proof against their own numbers, Lantern offers a free proof of concept: connect your ERP, build forecasts on real history, and see the impact before committing to anything. You can contact conor@lanternhq.com for more information.

Observations and Suggestions

• Inventory management / forecasting has historically been an issue for distributors. Perhaps with AI and the ability to “triangulate” additional insights, inventory, and purchasing, can be managed differently. The key is having a different mind-set. You need to be open to new ideas. If you are, I’d suggest taking Conor up on his offer … at least have a conversion.
• If you are a manufacturer who essentially sources everything, a major part of your business, if not all of it, is distribution. In this case, I’d also suggest reaching out. A conversation cannot hurt.
• And given what they are doing with wire / cable, if you are a wire / cable distributor, again, can a conversation hurt.

The key is that outreach needs to be two ways. The electrical distribution industry is known as a relationship industry. With only a couple of meetings a year, the opportunity to learn “new things” and meet “new people” gets diminished as companies need to revert to “smile / dial” sales (and the dial becomes email and Teams / Zoom.)

With budgeting season coming up and companies trying to determine their major 2027 initiatives as well as capital investments, now is the time to explore new ideas.