Auger

We had a great event last night to kick off our presence in Dallas. Two years in, we've built out our technical organization in Seattle. Now we're opening a Dallas office and building our customer-facing teams and part of our science org here. Besides the opportunity to grow our team of Augernauts from the amazing supply chain and operations talent in this city, this is the start of us building a community. Supply chain is a profession that runs on relationships and hard-won judgment, and very little of that gets written down. It lives in the people who move product, run plants, manage suppliers, and solve problems most of the world never sees. Dallas has a remarkable concentration of them. Thank you to everyone who came out. First of many. If you're in supply chain, operations, or procurement in DFW and want to connect, let us know. https://lnkd.in/gRJuqckW

Auger CEO Dave Clark, the former Amazon operations chief, told me yesterday that the decision to move the supply chain startup's HQ to Dallas — while maintaining and expanding a large engineering office in Bellevue — is about the specific talent needs of the company at this stage. Texas also happens to better suit him and Auger co-founder Leigh Anne Clark, and their family. He said it's not about taxes. But when I asked what he would say to people in the Seattle region who see another warning sign in the news, he didn't shoot down the premise. "If you're in that position," he said, "I think you're right to be worried." Read the rest of his comments in our story: https://lnkd.in/g3x4WWR9

Every decision in your supply chain runs on the demand forecast. And the forecast breaks down at the one time it matters most: when something has just changed the demand you are trying to predict. Most demand forecasts extend the past through a black box. That works as long as the future looks like the past, and you don't care why. Then a health study lands on a category, a competitor launches, a retailer resets its shelf, and the world is no longer the one the forecast was built on. The questions that matter now are how big the shift is and how long it lasts, and the black box can't answer either. A passing scare and a permanent shift look identical for the first few weeks, but they call for opposite responses: hold supply through the dip, or cut it before it strands inventory. Guess wrong and you over-cut a line that recovers, or overproduce into a decline that doesn't. Wait for the trend to settle the question, and the window to respond well has already closed. Here is why the forecast cannot help. It reads only its own past, sales and seasonality, so it never sees the world that actually moves demand and never estimates how demand responds to it. It projects the line; it does not model the cause. So it cannot size a genuinely new event or simulate how a plausible scenario plays out. A planner can guess, once, by intuition. Doing it properly, the moment the signal appears and across every line it touches, is beyond the tools on the desk. What is new is not that demand can be modeled. What is new is that the data and that expertise now sit in the same place: an agentic team that reads the outside signal, estimates how demand responds to it, and puts a size, a reach, and a duration on the shift the moment it appears. Because it models the cause rather than extends the past, it can simulate the scenarios that are actually likely, and feed the sized, time-bounded result into the sourcing, inventory, and distribution decisions that all run on it. This is the whole series in one decision. What to buy, what to make, what to hold, how to move it, each runs on the demand signal, and each is only as good as how fast you can size that signal and trust it. Knowing what you should know, in time to act, is the new definition of efficiency. That's Auger. #SupplyChain #DemandPlanning #AI #EfficiencyRedefined #Auger

Dave Clark helped build Amazon’s logistics machine. Now he and his wife, Leigh Anne, are back in Dallas trying to build a $100 billion company of their own. After two years in Seattle, the Clarks have returned to North Texas with their two sons—and they’re bringing the headquarters of their fast-growing tech startup, Auger, with them. The company, founded in 2024, is taking on a problem Dave knows particularly well after 23 years at Amazon: the messy, fragmented technology behind the world’s supply chains. Auger has already raised $150 million and landed some serious names as customers, including Meta, Kimberly-Clark, and Fanatics. But Dave’s ambition for the company is considerably bigger. By 2030, he wants half of the nation’s non-services GDP flowing through Auger’s platform. That’s a staggering target for a two-year-old startup. And now, the effort to get there will be headquartered in Dallas. I sat down with Dave and Leigh Anne to talk about why they came back, why they believe Dallas is the right place to build Auger, and what exactly their technology is trying to fix. We also got into what it’s like building a company together, Leigh Anne’s role in the business, their plans to reengage with the Dallas community, and why Dave believes the technology underpinning modern supply chains is badly overdue for an overhaul. There are some great quotes in this one—and an interesting look at what could become a major new player in Dallas’ tech scene. Read the full, exclusive story only on D CEO's site right now: https://lnkd.in/gptsDhaP #DallasBusiness #DallasTech #SupplyChain #ArtificialIntelligence #Amazon

Everyone is debating whether AI is worth the cost. Meanwhile the cheapest freight lane you booked this morning is silently running up a bigger bill somewhere else in the company. Freight gets bought the way it gets measured: on cost per unit moved. Pick the cheapest carrier, lane, and mode that still makes the delivery window, and the transportation line comes in under budget. But the cheapest lane is usually the slowest and the least reliable, and both cost money somewhere else. A slower lane keeps inventory in transit longer and forces more of it upstream to cover the wait. A less reliable one widens the range of arrival times. Safety stock is sized against that variability, so an unreliable lane quietly raises the buffer every node has to hold. The freight budget comes in under target, and inventory and service absorb the difference. Pricing that tradeoff would mean connecting each routing choice to the buffer it forces and the stockout it risks, across the network, for every shipment. Transportation cannot see those downstream costs. They land on other desks and in later weeks, so they never enter the lane decision. The expensive "cheapest" lane keeps getting booked. What is new is that the whole cost is now computable at the moment of the decision: an agentic team that chooses carrier, lane, mode, and timing against total delivered cost, the freight plus the inventory it ties up and the service it puts at risk, using each lane's reliability rather than its price alone, and books the move. That's Auger. Next: the signal underneath all of it. #efficiencyredefined #supplychain #logistics #ai #economics #auger

Dallas: Come meet Auger. AI, supply chain autonomy, and a little bit of trouble. Join us in Dallas for an evening with the founders, builders, and operators shaping the next era of supply chain autonomy. Grab a drink, meet the people behind Auger, and connect with others working at the edge of AI and the physical economy. For anyone looking for a little bit of trouble, we’re hiring across product, engineering, customer transformation, enterprise sales, and applied AI in Dallas and Bellevue. If you ship, sell, deploy, or model systems that move the world, we want to meet you. New grads ready to prove themselves are welcome too. Request your invite: https://lnkd.in/ePyhF4W3

Your network is overstocked and still running out. Both are true at the same time, and they share one cause. Every stocking location protects itself. A plant, a distribution center, a forward warehouse each holds enough safety stock to meet its own service target against its own swings in demand. On its own, each of those decisions is sound. The problem is that demand risks at different locations are not independent. When demand runs high in one region, it is often low in another. A shortage at one node tends to coincide with a surplus at another. Because these swings partly cancel out, the risk the firm faces across the whole network is smaller than the sum of the risks at each location. If every location stocks for its own worst case, the firm ends up holding enough for a day when every location has its worst case at once. That day essentially never comes. So the firm carries more inventory than it needs. And because that inventory is spread across locations that were each covering only themselves, it is still in the wrong place when a real shortfall hits. The firm holds too much overall and still runs short where it matters. Fixing this means sizing inventory to the firm's risk, not to the sum of local risks. That requires one current view of every node at once: what is on hand, what is in transit, what is inbound, and what demand is coming. It also requires knowing, when a shortfall appears, whether it is local or network-wide, because the two call for different responses. Assembling that view by hand takes longer than demand takes to change, so most firms do it once a planning cycle, if at all. In between, every location goes back to buffering against its own uncertainty. What is new is that this is now something you can compute continuously. An agentic team keeps one live model of position, in-transit, inbound supply, and demand across the network. It projects which nodes are about to run short and whether the shortfall is local or national. It sizes inventory to the firm's pooled risk rather than to the sum of independent buffers, decides where stock should move given where it is, where it is needed, and what moving it costs, and executes the transfers in minutes. That's Auger. Next: getting it there. #EfficiencyRedefined #SupplyChain #Inventory #AI #Economics #Auger

Everyone is asking whether AI is worth the cost. The sharper question is whether it can make a decision siloed teams never could, and cutting the products that quietly lose money is exactly that kind of decision. Every new SKU passes an easy test: at expected variable cost, it adds margin. So the portfolio only ever grows. A variant for one retailer, a seasonal pack, a slightly different size, each clears the bar on its own, and none is ever the one to cut. But that is not the whole cost. The true cost of a variant is what it does to the plant and the balance sheet: another changeover, a shorter run, more safety stock, the risk it ends up obsolete. Its true value is not its own margin but its margin net of the demand that would simply move to the product beside it on the shelf if it were gone. A SKU can be profitable on its own line and destroy value for the firm once you net out the manufacturing complexity it adds and the sales it takes from its siblings. And that's a complex calculation. The cost is absorbed into aggregate plant and inventory costs and never charged back to the SKU that caused it, and the substitution question needs a demand model, not a sales report. So the number that would justify cutting a variant remains invisible, and margin-negative SKUs continue to sit in the line, quietly eroding margin. What is new is that the true price is now computable: an agentic team that scores each variant against its full cost to make and serve, and against the demand that would substitute to its siblings if it were dropped, ranks the portfolio, and sizes the margin and working capital that rationalizing it would release. The portfolio decision that never got made becomes one you can run on demand. That's Auger. Next: the warehouse. #EfficiencyRedefined #SupplyChain #Manufacturing #AI #Economics  #Auger

Congratulations Alex Ceballos!