cubic: Product, team and how it was built

Started small, grew after funding · sold 11 months after launch

Its founders wrote some of the methods that the industry already used to run models cheaply, such as AWQ for 4-bit weights. They turned that research into a paid serving API. They proved the speed on a public benchmark and partnered with a cloud company that needed exactly that speed. Eleven months after its first public post, the partner agreed to buy it.

Started small, grew after funding · sold 37 months after launch

It aimed at regulated enterprises that needed AI agents on their own servers, with audit logs. The company said in 2025 that hundreds of organizations used it. Asana bought it in May 2026. Asana's quarterly report states the price.

Solo founder · 2 months after launch, revenue flat

AI coding agents can build a working page or deck, but in the founder's words their visual output often looks generic. Skillry sells finished visual styles, one task per Skill, shows real output before the sale, and adds new Skills almost every day under one subscription. When sales began on 28 September 2026, they held at about US$430 a day for 17 days.

Solo founder · 12 months after launch, revenue falling

It launched in October 2025. By then, Anthropic's Claude Code had made skills, slash commands and subagents a normal way to steer a coding agent. Most developers were still writing their own. A founder who presents themself as a CTO with 18 years of experience sold a tested set as a one-time purchase with a money-back promise. Verified revenue reached US$16,040 in March 2026 and US$82,195 in total, but it fell to under US$2,000 a month by September 2026.

Solo founder · 4 months after launch, revenue falling

Said over US$36K in a monthRecords US$39.6K by end of JulyHeld

It turned a new working habit into a native Mac app. The habit was running several AI coding agents at the same time. The app had voice control and a strong visual style, and it charged once instead of every month. The founder sold it to his own audience through daily live streams, X and a Discord community. Stripe recorded more than three quarters of the revenue in the first seven weeks.

Solo founder · 6 months after launch, revenue growing

The compression engine already existed as a free, open-source Python tool, but it was hard to install. Garm Lucassen packaged it into a desktop app. The app installs with a double click, connects the coding agents by itself, and removes its changes when it quits. The engine's creator suggested that he build it and keep the Headroom name. Developers who hit the usage limits of their AI plans pay a small monthly price to make those limits last longer.

Solo founder · 27 months after launch

The founder built it for his own inbox and published the code from the first day. He launched it on Product Hunt the day after Christmas 2023, where he says it finished first. The open-source community, two newsletters and the GitHub trending page brought the first thousands of users. According to the Startup Founder Stories summary of his Indie Hackers interview, he then narrowed the product to a niche when larger competitors arrived. The same summary reports that he reached US$10,000 in monthly recurring revenue about a year after the first customer.

Solo founder · 7 months after launch, revenue growing

It went where the users already were. It became an app in ChatGPT's app store and a plugin in Claude's directory and in Claude Code, with setups for Cursor and Codex. It did this in the months when these directories were new. It charges per tool call instead of a share of ad spend, so a small advertiser starts at US$49 a month. And it applied to the platforms' own API programs. The company says Google approved it in June 2026 and Amazon admitted it to its partner network.

Team of 2 to 10 · 23 months after launch, revenue growing

The first version was an AI copilot for DevOps engineers, who could build their own agents. The founders moved the product to non-technical teams that answer the same questions all day. They started every customer with one small use (repeated questions in Slack). They added a higher-priced plan when power users asked for more agents. The founder said in a 17 October 2025 interview, quoted by SoFarBot, that this took the company from US$100,000 to US$400,000 in annual recurring revenue.

Team of 2 to 10 · 27 months after launch

The founders had run conversion experiments at large companies, and the product removes the slowest step of that work. The Y Combinator listing says the chief executive ran experiments at Uber and the chief technology officer worked on Alexa AI. The company's launch post says a paywall change took at least 6 weeks, because it lives in native app code. Helium moves the paywall out of the app build.

Team of 2 to 10 · 33 months after launch, revenue falling

The product goes where the customers already are. A shop owner does not install a new app or learn accounting software. Instead, the owner writes into a chat. The founder wrote of having worked as a data scientist at a microloan company and of running a café. So the founder knew both the bookkeeping problem and the reason banks ask for records.

Solo founder · 19 months after launch, revenue growing

The founder, a digital nomad, had collected more than 1,000 receipts from many countries to track his own spending. He built the app for that use. It reads every line item instead of only the total, which matters for travellers with receipts in foreign scripts and currencies. Small, steady releases widened the market: 25 versions between 21 January and late October 2026. A translation of the app into 52 languages did the same, without a team.

Solo founder · 13 months after launch, revenue growing

A founder who had sold courses since 2022 built the email tool that he wanted for himself, and priced it by emails sent. His own newsletter was the first test of it, and the product now lets AI agents run an account.

Solo founder · 12 months after launch, revenue falling

The founder built a tool that he wanted for himself as a runner. He trained with a human coach from late July 2025 and stopped a marathon at kilometre 32. The product keeps what he valued in that coaching: varied sessions, pace ranges and a plan that follows what he runs.