What we do in one line
LeedAB is the operations layer SMBs install on their own machines — a roster of specialised AI agents that share one customer-specific brain and run without sending data to anyone's cloud.
The problem. Small and mid-sized businesses get the same pitch every quarter: "let AI run your ops." They try the chatbot, get a 30-day demo, and bounce. Why? Because the agents have no memory of their business, no shared context across departments, and they all need the customer's data shipped to a cloud.
The wedge. A LeedAB install drops a per-customer "brain" — a structured knowledge vault — onto the customer's own machine, then runs specialist agents on top of it: Ops, Procurement, Finance, CX, Sales, Marketing, plus a COO orchestrator that fields every message and delegates. Customer data never leaves their hardware.
The moat. Once a customer's brain is populated and their workflows are wired, the cost of switching isn't "buy a different chatbot" — it's "rebuild the substrate." Lock-in is structural, not contractual.
SMBs in regulated verticals (logistics, healthcare, legal, trades) are increasingly told they "can't put that data into ChatGPT." Local-first stops being a niche concern; it becomes a procurement requirement.
A 2026-spec Mac Mini runs Claude-class models well enough for ops tasks. The "must be cloud" argument is yielding — appliance shape becomes feasible inside the next 12 months.
Today's agent products treat memory as an afterthought. We treat it as the product — every agent reads from and writes to one structured per-customer brain. The agents are interchangeable; the brain compounds.
From bottom to top: substrate, license, manifest, agents, channels. Each layer is small, well-defined, and replaceable. The full stack runs on a customer's Mac Mini today (still software-only); appliance shape is on the 12-month horizon.
31 specialist agents shipped today across executive (COO, Concierge), growth (Marketing, Content, Campaigns, Email, Brand, Video), build (Product, Engineering, Frontend, Backend, DevOps), automation, research, people, ops + supply, finance + investor + data, support (CX), legal + compliance, and customer lifecycle (Solutions, Onboarding, Implementation, Champion, Delivery).
Ex-Microsoft Azure PM (Sydney, FastTrack + Red Flag, 2022 – 2025). Masters in AI at UTS. Masters in Cybersecurity (Macquarie). Native Arabic + native English. Builds the substrate.
Long-form software engineer. Ships the production console + license backend + CI surface. The "this needs to be a product, not a demo" voice.
Engineering hire joining Feb 2026. Onboarding patterns being formalised against Muiez's join experience.