Building autonomous AI agents. Notes from the team building Gravity. AI workflows, the future of recurring work, and what we learn along the way.
As of mid-2026, there is no single AI agent law to comply with; there is a layered landscape you have to navigate at once. The EU AI Act is in phased enforcement, several US states have their own AI rules, the NIST…
Read post →
Prompt injection is the attack where text fed to a model is written to override its real instructions, and you defend against it by assuming no single control will stop it. There is no filter that reliably separates…
Read post →
An AI agent implementation goes right when you start with one narrow, well-defined workflow, give the agent least-privilege access to the systems it needs, keep a human approving consequential actions, and measure…
Read post →
An AI agent can assemble your weekly status report by pulling from the tools where work actually happens. It reads the project tracker, the git history, the CRM, the shared docs, and the team chat, then summarizes…
Read post →
Every team has a question that gets asked once a week and answered the same way every time. "What is the staging database URL?" "Who approves a refund over five hundred dollars?" "Did we ever decide on the new logo…
Read post →
An AI agent can schedule meetings on Google Calendar end to end. It reads the calendars of everyone invited, finds times when all of them are genuinely free, respects each person's working hours, buffers, and focus…
Read post →
Most teams treat content as something you switch on at launch, alongside the press push and the paid spend. It feels logical: why write about a product nobody can buy yet? The trouble is that content does not behave…
Read post →
WWDC is Apple's developer conference, and in recent years it has become one of the most-watched moments in consumer AI, because Apple's choices shape what billions of phones can do by default. The interesting…
Read post →
For most of the early agent era, the pitch was breadth. The promise was a single general-purpose agent that could plan, browse, write code, and handle whatever you threw at it. That vision is still alluring, but in…
Read post →
We have published more than 500 blog posts before launching our product, and the traffic is still a trickle. That is the honest headline, and we are writing it down on purpose. It is tempting for an early-stage…
Read post →
When you open the Gravity catalog, you see a list of agents that each do a specific job, and you can run any of them in about 60 seconds by describing what you need. What you do not see is the work that decided which…
Read post →
Most launch announcements arrive after the fact, polished and certain. This is the opposite kind of post. Gravity is still pre-launch, on a waitlist, and this is a candid look at exactly where things stand: what is…
Read post →
Google's Gemini line has moved steadily from a chat-and-completion model toward a foundation for agents: software that does not just answer but takes actions, calls tools, and works through a task. The direction has…
Read post →
"Are AI agents overhyped?" is the wrong question if you expect a yes or no. By mid-2026 the answer is plainly both. The technology is real, it does useful work every day, and the curve of what it can do has bent…
Read post →
By 2026, AI agents are no longer a question of whether but of how, and the how looks very different depending on the size of the buyer. A 40,000-person enterprise and a 12-person agency both want agents to do real…
Read post →
To build a research agent, you set up one repeatable pattern: gather information from defined sources, verify it, then synthesize a summary you can trust. In practice that means deciding the exact research question,…
Read post →
There is no single best AI agent for sales teams. The right one depends on where your pipeline actually leaks: top-of-funnel sourcing, follow-up that slips on a busy day, or a CRM that nobody keeps clean. So instead…
Read post →
The best AI agent for marketers depends on where your work actually slows down. The content usually gets written; what slips is the rhythm around it: repurposing last week's post into channel-specific versions,…
Read post →
AI agent platforms diverge on six dimensions that decide fit: how you tell the agent what to do, how you pay, whether the platform is no-code or code-first, how it connects to your tools, whether it supports human…
Read post →
Yes, an AI agent can run your post-webinar follow-up end to end: it pulls the registrant and attendance lists from your webinar tool, splits them into attended and no-show segments (and sorts attendees by engagement…
Read post →
Yes, an AI agent can turn a finished podcast episode into a complete set of show notes. Hand it the transcript, or the audio with timestamps, and it returns a tight episode summary, timestamped chapters, key…
Read post →
Yes, an AI agent can analyze NPS survey responses end to end. It computes the score and splits respondents into promoters, passives, and detractors, then reads every open-text comment, groups the verbatims into…
Read post →
Yes, an AI agent can manage a content calendar end to end: tracking every planned piece and its status, owner, and deadline across the tools the calendar lives in, flagging any upcoming week with no draft, chasing…
Read post →
An AI agent can monitor churn risk by watching the signals that precede a cancellation across every system they appear in, scoring each account, and surfacing the ones trending toward churn early enough to act. It is…
Read post →
The right way to buy an AI agent platform in 2026 is to start from the job you want done, score every vendor on the same short list of criteria, and prove it with a small paid pilot before you sign anything. This…
Read post →
SaaS teams run on recurring work: onboarding new accounts, watching usage for churn signals, triaging tickets, chasing failed payments, and turning product data into reports. AI agents are built for exactly this…
Read post →
Secrets rotation is the practice of replacing an AI agent's credentials, the API keys and tokens it uses to reach models, tools, and data, with fresh ones on a schedule, and retiring the old ones. It sounds like…
Read post →
PII redaction for an AI agent is the discipline of finding personal data in everything the agent touches and stripping or masking it before it reaches a place where it should not live: the model prompt, an outbound…
Read post →
A compliance audit for an AI agent asks one blunt question over and over: can you prove it? Not "do you have access controls" but "show me who could reach this agent last quarter." Not "do you handle data carefully"…
Read post →
Sierra is one of the most credible names in enterprise customer experience AI as of 2026. It was founded in 2023 by Bret Taylor, a former co-CEO of Salesforce and chair of the OpenAI board, and Clay Bavor, a former…
Read post →