Keith McAleer · San Francisco
Currently exploring agentic marketing operating systems and orchestration.
A window into things I've built: products, platforms, talks and games. See the work
How the work connects
The same loop runs through a campaign, a university platform and a desktop agent. What has changed is how much of it I can now build myself.
Who is this for, and what are they actually trying to get done?
See: Narrow ICPs
Positioning and narrative clear enough to survive being retold.
The platform, CRM, analytics or agent that gets the story to the people it's for.
See: BEGIN
Did people use it? Did it pay back? Then do the next loop.
AI and product run underneath every stage: as a research tool, as a way to prototype the idea, and more and more as the thing being built.
Selected work
Three projects, each showing a different part of the same job.
01 · Desktop AI agent
Shipped · free beta for Windows and macOS
One voice in front of the AI systems you already use.
A voice-first desktop assistant and orchestrator. Neko answers out loud, keeps context for each project and client, runs local apps, and hands coding work to Claude Code or Codex in the background, inside safety boundaries enforced in code.
Agents
Tools by area
02 · Agentic marketing workspace
In development · not yet deployed
Making it easy for a whole community to contribute to its own marketing.
A multi-agent workspace that combines structured contribution forms, reusable brand context and agentic workflows. Faculty, staff and students can hand over stories and campaign inputs with little effort, and a small marketing team can turn them into useful content.
03 · Discovery platform
Live · relaunch in preview
Turning institutional complexity into a navigable product.
A structured discovery layer across 155 Berkeley entrepreneurship and innovation resources, organized by stage, audience and category. A student, researcher or founder can find where to go next without first decoding the university.
Things I build
Independent products and experiments, mostly built with AI coding agents. Some are tools I use every day. Some exist to find out what's possible.
iOS app · awaiting App Store release
A Korean flashcard app you answer out loud. On-device speech recognition grades each answer, and spaced repetition brings missed cards back. 55 decks run from A1 to B1.
Live · private workspace
A workspace where founders grade hypotheses, experiments and decisions on an explicit evidence ladder. The AI assistant proposes structured changes that the founder has to approve.
Experiments
Browser games built as experiments in game feel and in what AI agents can carry end to end. Rune Rush, a first-person wizard arena with boss waves and controller support, was built through Neko.
Live
A small app studio site with the publishing, policy and support pages that apps in the stores require. It publishes Kkachi.
Ideas I'm working through
Two talks from October 2026. The short version: AI made average content free, so point of view, proof and structure matter more than ever.
3 October 2026 · BCS AI + Recruiting Summit
Berkeley's innovation ecosystem, then six popular claims about marketing in the AI era, each tested against current research and given a verdict.
October 2026 · UDAX Berkeley Accelerator, Cohort 2
A practical framework for founders: positioning, story, brand, when to invest in marketing, and the mistakes early-stage teams make most often.
About
I started in communications, and over fifteen years I kept moving underneath the work: websites, then analytics, CRM, UX, automation, software, and now AI agents. Each time, the most useful answer to a marketing problem was a better system, so I learned to build them.
Berkeley's entrepreneurship center gave me founders, engineers and faculty to learn from, a federated organization to make sense of, and permission to experiment. I still write, design, analyze and ship.
Chancellor's Outstanding Staff Award, UC Berkeley, 2022.
More about how I workI'm based in San Francisco. The easiest way to reach me is email or LinkedIn.
