Latest from the build
The CoMavenAI blog
Notes from
the build.
What we're building. What we're learning. Practical perspectives from engineers working alongside real teams.
Inside the work
Embedded engineering. Applied AI. Lessons from production.
Explore the thinking
Find your next useful read.
20 articles and counting
Who owns an AI system after launch?
A production AI system needs named owners for the workflow, technical operation, security, review, and the exceptions between them.
In the Arena #8: two of 66 runs caught the $25.7 million
The best agent finished about a quarter of the finance assignments. Four reads on where agent work actually breaks, and where it is already cheap.
Build, buy, or automate inside the stack you already have?
A neutral way to choose among an existing product, an integration, workflow automation, and a custom build.
Get new posts by email
When we publish, you get it. No drip sequences, no sales follow-ups. Just the articles. Unsubscribe anytime.
Loading the form… You can also email contact@comavenai.com with the subject “Subscribe”.
In the Arena #7: same model, twice the bill
The software around a model can double what it costs to run. Four reads on what AI really costs, who is already using it, and what agents do when nobody checks.
Your AI team has a roadmap. Give them room to deliver.
Your internal team knows the business. A useful engineering partner takes responsibility for a defined piece of work and helps the next priority move forward.
Map the handoff, not the screen: where workflow automation pays
The expensive work often lives between screens and systems. Map the handoff, its exceptions, its owner, and proof that it finished.
In the Arena #6: your business isn't the frontier
Sam Altman and Dario Amodei want to pace the frontier. Your next useful software project needs the right model, clear rules, and economics that work.
8 questions I'd ask before hiring an AI RevOps consultant
The 8 questions I would bring to an AI RevOps conversation, from a working system and measurable outcomes to ownership and the people behind the build.
How to choose your first AI workflow
A five-factor scorecard for ranking real AI opportunities by frequency, cost, input stability, outcome, and exception risk.
In the Arena #5: when should the agent ask a human?
Where AI tools become operational change, when agents should hand work back to a person, and how to check whether one recurring task is worth a small test.
What could AI give you back? Start with one workflow.
A few hours back can matter to a one-person business, a nonprofit, or a busy team. Start with one recurring task and explore what a useful change would take.
Five AI projects we tell companies not to build
The fastest way to find a strong AI project is to get honest about the ideas that should not make it into the build queue.
In the Arena #4: ask for the twentieth run
The agent that passes once and fails the other nineteen, what companies actually bought in July, the one metric that survives cheap code, and the first of 33 questions executives asked.
What forward-deployed engineering really means
Being in your Slack and your Jira is table stakes. The real work is sitting with the people who actually know the business - and building with them, not around them.
In the Arena #3: the model is the cheap part
The build that beat the buy, five verbs that moved a benchmark 59 points, who is actually shipping with AI, and the agent that found the answer key.
In the Arena #2: two agents are not two opinions
Correlated blind spots in agent swarms, the OpenAI number worth reading, Agent Plugins, and who funds the compute.
In the Arena #1: what small teams actually do with AI
Small teams stretching jobs with AI, the sixteen agent ceiling, and the EU's new disclosure rules.
Introducing CoMavenAI: your vision, engineered
StratusPointIT launches a new AI subsidiary and Queen City AI becomes part of it. What it means, and why we built it this way.
What is a Forward Deployed Engineer?
The AI role everyone is suddenly hiring. What FDEs actually do, and how the mid market gets one.
We run our own outbound on an AI agent swarm
The agent system behind the Growth Engine, the numbers it produced, and what it taught us.