Building autonomous AI agents. Notes from the team building Gravity. AI workflows, the future of recurring work, and what we learn along the way.
If you searched "Gravity vs Fixie.ai" expecting two head-to-head agent platforms, the honest first thing to say is that Fixie.ai is not quite the product it was pitched as a couple of years ago. The team that built…
Read post →
AutoGen is Microsoft's open-source framework for building multi-agent systems in code, and it has been one of the most influential projects in the space, the place a lot of engineers first learned to make several…
Read post →
This is an unusual comparison, because by 2026 Adept is not really a product you can choose. Adept was one of the most admired startups in the agent space: its ACT-1 demos, where an AI operated real software the way…
Read post →
The honest answer to "what is the best AI agent platform for startups" is that it depends on your team and your stage, and anyone who hands you a single ranking is selling something. We build Gravity as a startup…
Read post →
A solopreneur is not a small startup. A startup is a team trying to scale; a solopreneur is one person trying not to drown. The whole game is leverage: how do you run sales, delivery, finance, and marketing alone…
Read post →
Most platforms that let you build something to sell take the lion's share and pay you once. Gravity is built the other way around. A builder earns 21.27 percent of every run their agent completes, as profit, for as…
Read post →
Halfway through 2026, AI agents are everywhere in conversation and still mostly absent from production. That tension is the whole story. The hype says agents are running companies; the data says most are running in…
Read post →
OpenAI's Codex is a coding agent, so it is tempting for anyone who is not building developer tools to skip it. That would be a mistake. Codex is one of the clearest production examples of how to build an agent that…
Read post →
Gravity and Dust both live under the AI agent banner, but they are built for different jobs. Dust is an enterprise platform for building AI assistants grounded in your company's data, the documents, conversations,…
Read post →
Most agent demos work. That is the problem. A demo is one run, on a clean input, with the builder watching. The moment an agent goes live it meets empty fields, rate limits, contradictory instructions, and the one…
Read post →
Every few months Google runs a broad core update, the search world panics for two weeks, and then everyone goes back to writing. A Q2 2026 core update is no different in kind, only in timing. This analysis cuts past…
Read post →
Reviewed against Gartner's latest published research as of July 15, 2026.
Read post →
Enterprise adoption of AI agents in 2026 is not a single story, it is a pattern: agents enter through a few high-volume functions, run as pilots before they run as production, and live or die on governance and proven…
Read post →
When people ask which model to build an agent on, they usually want a ranking. The more useful question is which capabilities an agent actually depends on, and how a given model delivers them. Claude 4 is a strong…
Read post →
The EU AI Act is the first comprehensive AI law from a major jurisdiction, and in 2026 its obligations are no longer theoretical, they are phasing in on a fixed schedule. For anyone building or deploying AI agents…
Read post →
Founders love to publish waitlist numbers because they look like traction and demand nothing in return. I want to do the less satisfying thing: explain what an early, pre-launch waitlist genuinely tells you, what it…
Read post →
I have started over enough times to be suspicious of momentum that I cannot measure. So at the end of the first stretch of building Gravity in the open, I want to write down what is actually true rather than what…
Read post →
Publishing roughly ten posts a day for a month is an unusual thing to do, and I went in expecting to learn about SEO. I came out having learned more about the difference between producing a lot and saying something,…
Read post →
I have watched a hype cycle from the inside three times, and the first half of 2026 had the unmistakable texture of one turning. Not a crash, a sorting. The capital was still there, the demos were still dazzling, and…
Read post →
Marketplaces are the hardest business model I know of, which is exactly why I chose one. The easy models are easy because they are crowded; the hard ones, when they work, are defensible because almost nobody gets the…
Read post →
I have pitched ideas that were technically better than the thing that got approved, and lost. The lesson stuck: the quality of an agent project rarely decides whether it gets funded. The quality of the buy-in does. A…
Read post →
Most teams do not migrate to AI agents because a vendor sold them on it. They migrate because a Zap broke on an edge case for the third time, or because a Make scenario grew into a 22-step chain that nobody dares…
Read post →
"How long will this take?" is the first question every buyer asks and the one most vendors answer badly. The honest answer is that it depends on which of four very different things you are actually building. A…
Read post →
An executive does not read a business case to learn. They read it to decide whether to bet a slice of budget and reputation on you being right. Everything in the document either reduces their uncertainty or wastes…
Read post →
Choosing an AI agent platform looks easy until you have to defend the decision to a CFO, a security review, and a procurement officer simultaneously. The marketing pages are interchangeable. The pricing is opaque…
Read post →
Agent updates fail badly when they are in-place. A prompt edit lands mid-run and the second half of a conversation no longer matches the first. A new model lands and a tool-call signature shifts. An index rebuild…
Read post →
ROI calculations for AI agents fall into two categories: defensible numbers backed by measurement, and made-up numbers backed by vendor claims. The CFO can tell the difference. This guide is the defensible version:…
Read post →
An agent PoC succeeds when the go-no-go decision is obvious within 6 weeks. It fails when scope creeps, baseline is missing, or no one is responsible for the call. The 25-item checklist below covers what to confirm…
Read post →
This is the RFP template I send when buyers ask "give me the question list". Sixty questions across six sections, with a 1-to-5 scoring rubric and walk-away criteria. The template assumes enterprise procurement;…
Read post →
A PoC tells you the technology works. A pilot tells you the deployment works. Most teams skip the pilot because the PoC succeeded; then production hits real volume, real users, and real operational concerns, and the…
Read post →