Operations, logistics and trades
Software for the daily work of transport and logistics companies, factories, metal service centers, construction estimators, tradespeople, stores, procurement and HR teams: routes, production plans, quotes, supplier checks and work instructions.
Operations, logistics and trades holds 31 of the 603 stories on Aiolum: 3 run by a solo founder and 28 by a team of 2 to 10. Together they report US$32,340 in verified outcomes; the largest is Tutti, US$2.7K in verified revenue in 30 days, October 2026. 1 headline number carries evidence grade A (disclosed or verified), 0 grade B (independent press) and 0 grade C (founder claim), 10 are grade D (reported, not verified), and 20 have no revenue figure (grade E). The median story reached its headline number 17 months after launch. The founders are in 15 countries, most often United States and Germany. The most common main channel is cold outreach (3 of 31). The tools named most often in their sources are Google Analytics (15), Cloudflare (14) and Next.js (11).
Verified, number by number
Evidence gradesA disclosed or payment-verifiedB independent pressC the founder's own figuresD reported, not verifiedE no revenue figureHow we verify
Filter: team, grade, outcome and more
More filters: main channel, size and time
The revenue trend compares the latest months in the sources with the months before them. Falling means revenue is about 15% or more below the best recent months, or the sources report a fall. The headline number is the yearly figure or the sale value on each card.
31 of 603 stories shown
Tradespeople already use WhatsApp all day on the job site, and many write quotes by hand or in a spreadsheet. Tutti puts the office work into that chat, so the customer has nothing new to install or learn. It sends messages through Meta's official WhatsApp Business Platform, which lets one founder run the service without a phone app as the main product.
5 grade-A facts · 4 that failed before itIt sold to large companies from the first months, together with a network of human mentors, and an investor reported that it made €30,000 a month within a few months of launch. The founders then turned the mentoring business into a platform built on skills data, raised a €5.8M Series A in 2025, and the managing director of Techstars Torino reported €3.6M in annual recurring revenue in February 2026.
Every number graded · 4 that failed before itLarge companies run many HR systems that do not share data, so a single case such as a new hire passes through several of them by hand. Ahron does not replace those systems; it connects to them, measures where cases wait, and places agents there. The founders had sold an earlier HR software company, and the product is built on Celonis, a Munich process-mining company whose partnership, the founder says, brings in requests from around the world.
Every number graded · 4 that failed before itThe founders came from a company that did outsourced purchasing for other firms, so they knew which purchasing tasks were still done by hand in email and spreadsheets. Soource turned those tasks into agent workflows for large companies. One article from the South Summit Madrid 2026 conference reported 800K in annual recurring revenue and 58 enterprise customers, without a currency and without saying who gave the figure.
Every number graded · 4 that failed before itFrench renovation quotes depend on subsidy and tax rules that change every few months, and the small firm that installs the equipment bears the risk of a wrong file. Argile puts those rules and the building calculation methods into the software, so a salesperson can give a detailed quote within an hour of a visit instead of weeks later.
Every number graded · 3 that failed before itFood distributors still take many orders by email, text message and phone and type them into old ERP systems by hand. The two founders worked inside a Boston seafood distributor for months to map that work before they built the product. Then they sold it themselves to large distributors: an a16z speedrun newsletter of January 2026 says one enterprise deal with a publicly traded distributor took their booked revenue past US$500,000.
Every number graded · 4 that failed before itThe founders say that planners in waste depots drag jobs onto trucks by hand every morning, that the plan is wrong within hours, and that companies cannot hire planners fast enough. They had sold an ERP to waste operators since 2022 and learned that large operators would not replace their ERP, so the new product sits on top of the ERP instead of replacing it. In the April 2026 launch post the founders said that annual recurring revenue was US$496K.
3 grade-A facts · 4 that failed before itA slitting plan decides how each wide coil is cut into narrow strips for many orders at once, and the company's own pages say most planners still build it by hand in a spreadsheet. A co-founder came from several years of work in a service center, so the product was built around the rules of that floor (grades, gauges, knife setups, coil inner and outer diameters) instead of a general factory model. The company charges a share of the savings it measures, so a buyer pays only when scrap goes down.
Every number graded · 4 that failed before itTwo founders who had run a car rental side business saw that fleet companies kept their planning, tracking and invoicing in separate tools and lost money in the gaps between them. They sold one control center that joins operational and financial data, and in May 2025 they wrote that revenue had reached US$200,000 a year, all from word of mouth.
1 grade-A fact · 3 that failed before itThe founder said in November 2025 that he passed US$100,000 in annual recurring revenue alone, without outside money. The trade press reported how he got there: he found one Bay Area retailer with seven stores, watched how its associates used the app, and built what they asked for. Each new feature tied the product deeper into the store's own data, such as the stock in the Ace Hardware warehouses.
Every number graded · 5 that failed before itPricing a tender is mostly reading: estimators spend days moving numbers from PDFs and spreadsheets into their own price build-ups before they can think about the price. The product starts from the documents the contractor already has, so a customer needs no connection to other company software. UpNext World reported in March 2026 that paying customers in Ireland gave it €5,000 in monthly recurring revenue.
Every number graded · 4 that failed before itGerman tradespeople must soon send electronic invoices to business customers, and many still write quotes in Word or Excel in the evening. AngebotFIX lets them dictate the quote on site and sells this at a lower price than the established trade software, with the founder answering support questions himself by WhatsApp.
1 grade-A fact · 2 that failed before itThe founder is a search-and-rescue dog handler who saw that teams still kept paper logs. Professional dog units must prove training for certification, audits and court, so a structured record with timestamps and edit history has a clear buyer. The AI summaries turn those records into advice without extra work for the handler.
Every number graded · 3 that failed before itBrazilian import costs are hard to predict: the co-founder says that taxes, freight and insurance are about half of the cost of an imported product, and more after a mistake. Many importers still calculate this in old spreadsheets. HeyShip turns the invoice into a cost estimate in minutes and adds warnings about controls and permits.
Every number graded · 4 that failed before itThe founder spent twenty years in supply-chain and production roles and saw new workers struggle to learn from scattered documents while experienced operators retired. The product records what experts do and turns it into guidance that a new worker can follow in their own language.
Every number graded · 3 that failed before itDistributors of plumbing, electrical and hardware goods keep tens of thousands of items in old ERP systems and do much of the reorder and quote work by hand. Lark sells agents that sit on top of the ERP instead of replacing it, so the customer does not have to migrate. The one published customer, Interline Hardware, says the reorder list now takes about 30 minutes a day instead of three hours.
Every number graded · 3 that failed before itThe founders worked as dispatchers at home-service companies before they built it, and they say dispatch decides how much a shop earns but runs on gut feeling. The software assigns jobs by skill, location, availability and close rate instead of by who is free.
Every number graded · 3 that failed before itMost development projects go through two or three rounds of city comments and resubmittals, and the founders say each round adds months. Many of the checks are the same on every project, so software can run them first and leave the judgment calls to the human reviewer.
Every number graded · 3 that failed before itTwo employees of a Cologne e-commerce agency saw that shop search was one of the weakest parts of their clients' shops. They built the first versions inside the agency, tested them with agency clients, and took the first customers from the agency's network before they spun the product out as its own company.
Every number graded · 3 that failed before itOne of the founders had a clothing store that kept receiving returns because of the wrong size. Most sizing tools of the time asked for height and weight. Fitit compares garment with garment instead, because, in the founders' words to El Observador, how a person wears clothes says more than the body measurements. Every store that joins adds wardrobes to the shared pool, which helps the next store.
Every number graded · 2 that failed before itThe CEO led analytics and sales teams for enterprise software in Japan and Asia for about two decades, he said in an interview, and saw that ERP systems record what happened but do not say what to do next under current constraints, so planners decide in weekly spreadsheet cycles. The co-founders built a layer above the ERP that keeps the evidence, the options, the approval and the outcome of each decision together.
Every number graded · 3 that failed before itThe founders saw that people describe a symptom ("the water does not drain"), not a job, so technicians come without knowing what to fix or what to charge. A language model that asks a few questions turns the symptom into a job that can be priced. They tested the idea with a web app that cost under US$100, as they told La Nación, and ran about 18 months on their own money before an angel round.
Every number graded · 4 that failed before itThe founders knew the trade before they wrote code. Antler wrote that the chief executive had co-founded an AI start-up for heating and cooling control, and that the commercial co-founder's real estate network had proved key to the early traction. Antler counted more than 300 hours of shadowing technicians by March 2025; a 2026 job advert says more than 600. The product then grew from a technician app into the whole help desk, where an FM company spends most of its admin time.
2 grade-A facts · 3 that failed before itFreight forwarding runs on a large volume of emails, PDFs and spreadsheets that people copy by hand between systems. The two founders, graduates of the Technical University of Munich, built agents for that one industry, with a record of why each decision was made, and found investors among former managers of large forwarders.
Every number graded · 3 that failed before itAt a distributor, inside sales reps type quotes and orders by hand, and the founder writes that growth in his family's business was limited by how fast it could hire and train them. Requests arrive as emails, PDFs and customer shorthand; the company says its AI maps that shorthand to exact catalog items. The price itself comes from rules that the distributor sets, so a person stays in control of what is sent.
Every number graded · 3 that failed before itMistakes in drawings, such as a door below the minimum width in the code or a duct that runs through a beam, are often found only on site, where fixing them means rework and delay. People do not check hundreds or thousands of pages against each other well. The founder knew the trade: by his own account he ran an engineering firm before and his father is a structural engineer. The company made the check something a firm can buy without a sales call.
Every number graded · 3 that failed before itAntler wrote in October 2024 that trades businesses often spend up to 3.5 hours per lead on surveying, scoping and preparing quotes. The founders started with one job, quotes for painters and decorators sent through WhatsApp, which tradespeople already use every day. When customers connected their accounting and email systems, the same agents took on more of the office work.
1 grade-A fact · 3 that failed before itA bid for a large building comes with hundreds of pages, and a missed line in a specification can cost the contractor money after it wins. The two founders first built a general AI tool for drafting business documents in 2024 and 2025, then focused it on preconstruction, where one firm's old bids and contracts can be reused on every new pursuit.
Every number graded · 3 that failed before itIt chose one small, repeated and rule-heavy task, the writing of customs entries, at a time when new US tariffs made that task larger and harder. The AI makes a draft and the licensed broker keeps the final check, which made the product easy to accept. A larger trade-software company bought it to complete its own customs workflow.
2 grade-A facts · 3 that failed before itRuhaab Markas started it after he saw how hard the back-office work was in a residential construction company he had invested in. He and John Baker, who had both worked on AI products at large technology companies, built it in about 18 months and then sold the company to Buildertrend, the construction management software company whose data and customers they wanted to work with.
Every number graded · 3 that failed before itIn February 2025 the two founders sold blueprint reading for many trades, with flooring suppliers as the first clients. By July 2025 they had trained a model for one trade, Division 8, where one building can have hundreds of doors, each with its own hardware set, and where the estimator otherwise types every line by hand into the quoting system.
Every number graded · 3 that failed before itThe next verified one-person AI business goes here. Same structure, same evidence grades.
Browse by what the product does
Patterns to copy
Every number on these cards carries an evidence grade. How Aiolum verifies explains the evidence grades and the five parts of each story.