Cerevisor is the AI controller for your work.

Why Cerevisor?

Cere- from cerebrum: cognition, strategy, the part that plans. -visor for vision, a lens that filters the noise, and the instinct to supervise the work rather than hand it off blind.

Cerevisor is the AI controller for business workers and power users: the place where the AI tools you already use work as one supervised team, on any model, with every run recorded on your machine. Your keys and your data stay there.

Why I built Cerevisor

My name is Kinan, and I founded Cerevisor. Almost two decades across strategy, marketing and automation led to my most recent role, Director of AI Strategy and Transformation at Cambium Networks, where the November 2022 GPT moment became enterprise-wide adoption, with 80% weekly active usage among licensed Claude users.

That work was hands-on: forward-deployed engineering across departments, an AI-powered go-to-market platform, a multilingual competitive positioning agent, and monthly company-wide AI training.

Curiosity about the human mind and the way we think has shaped my reading since 2006, alongside a passion for technology, space and human exploration. AI joined that list in 2012, and it has been the centre of it ever since.

Cerevisor started as my own tool for exploring the AI frontier. It is always ready to share, so anyone can put the latest in AI to work without being an AI engineer.

The widening AI adoption gap matters deeply to me, so Cerevisor keeps getting easier to use, for every kind of user.

• Director, AI Strategy and Transformation, Cambium Networks
• 80% weekly active usage among licensed Claude users
• Star of Cambium Award, 2022 and 2024
• M.Res in Applied Economics and Business, Universitat Autònoma de Barcelona
• English, Arabic and Spanish
• Based in Barcelona

Kinan Hamwi on LinkedIn

Built from daily practice

Shaped by multi-agent workflows that run every day, and by what holds up in real work.

• Built around real work, on multi-agent workflows that run every day
• Close to the model ecosystem: new models, agent tools and MCP servers get tried the week they ship
• Clear about what agents can and cannot be trusted with yet, with guardrails built for the gap
• Local-first is a principle: your prompts, keys, and runs stay on your machine
• One test for every feature: can you rerun the work next week and trust the result?

Mission

Make AI agents something a person can supervise: a check before every run, humans in the loop where judgment matters, keys and data on the machine, and every run on the record.

Vision

Everyone who works with AI runs a team they can see, steer, and trust, with their expertise amplified by AI.

Design Principles

The values baked into Cerevisor

Local-First by Default

Prompts, runs, MCP traffic, and provider keys stay on your machine. Nothing flows back to Cerevisor.

Provider-Neutral

Claude Code, Codex, Cursor, Antigravity, Grok Build, Anthropic, Gemini, OpenAI-compatible endpoints, and local Ollama. Mix providers per agent in one workflow.

Audit-Ready by Design

Every run writes a log on your machine, so 'why did the agent do that?' is answered from data.

Human-in-the-Loop

Approval checkpoints, escalation paths, and review steps belong on the canvas wherever judgment matters.

Composable Files

Workflows and worlds are plain files (.cerevisor and .cerevisor-world) you own, version, share, and re-run.

Visual and Open

A canvas and a chat builder for people who think in workflows, with room for code when you want it.

What Cerevisor Gives You

Why power users pick it over juggling separate AI tools

• Runs locally on your machine, and usage data stays there
• Any model per agent: Claude Code, Codex, Cursor, Gemini, API models, and local Ollama
• A local record of every run: approvals, tool calls, files touched, and cost
• MCP tools out of the box, no glue code to maintain
• Plain-file workflows and worlds you can version with git
• Free to start with a 7-day trial of everything, no card required