Operations people who build with AI.
inocta.io is an operations practice in Toronto and Montréal, in English and French. We read how a company runs, write it down so people and machines can both follow it, and put AI where the work can carry it.
Put AI inside real operations, and make it hold over time.
We exist to close the gap between what AI does in a demo and what it delivers on a Tuesday morning, in a real company, on work that matters.
Every company already has both parts: the intelligence and the operations. What is usually missing is the link between them, because the work was never clearly documented. Building that link is our job.
Intelligence is the machine layer: code, data, systems, automation, an intelligence quotient (IQ) that scales without tiring.
Operations is the human layer: process, judgment, an emotional quotient (EQ) that builds the path forward.
Humans and machines operating the business together.
We want to work at the heart of the question every company will soon have to answer: how do humans and machines share the work, and who decides what?
For us, this is not a prediction. It is a practice to build. The tools, the methods and the processes that will frame that collaboration have to be designed by someone, and we would rather build them than wait for them to take hold.
Today that work is done with AI. In time it will be done with machines that move. We are not building robots and we will not pretend otherwise. What we are building is the layer underneath: how the operation runs today, the written process, the boundaries on what a machine may decide alone. That layer is what makes the next one possible, and it is the same work either way.
It is why the full legal name has always been Inocta Robotic Technologies. We carry the vision in the name.
What goes in decides what comes out.
Whatever enters a process determines what leaves it, and that is as true of an AI as it is of a person. The name says so.
In, what enters the process
Point an AI at work nobody has written down and it will automate the confusion faithfully. So we read the operation first, before any technology is chosen.
You can't automate what you don't understand.Octa, and what eight carries
The eight-point star carried on the flag of the Marinid and Saadi Moroccan dynasties. Eight is the number that carries prosperity, and this company exists to create value.
The mark you see on every page.IO, input and output
At the founding it meant input and output: the digital foundation, and the way a model is trained on what it is given. It grew a second reading that became the primary one, and it is the promise the company is measured against.
The Intelligence of Operations.We built the operations before we built the company.
inocta.io was founded by Mo Kahlain after selling, in 2024, the company he had built, having earlier been part of another exit as a shareholder. Years spent building companies and working inside large corporations taught him how operations scale. Writing Adapting to AGI, then testing new technology ventures, showed him how quickly AI was advancing and how little most businesses understood about what it needs to deliver real value.
So he went back into the field and did the work directly with manufacturers and distributors, hospitality, architecture and construction, media, finance, e-commerce, education, and the agencies that serve them. The lesson was consistent: businesses were trying to implement AI, but the knowledge needed to make it useful was not documented, structured or accessible. AI was being pointed at work the business itself could not clearly explain. The problem was rarely the model. The operational foundation was missing.
inocta.io was built from this work to create that foundation: map the operation, structure its knowledge, then deploy and manage AI against it.
We are based in Toronto and Montréal, we work in English and French, and every document we produce exists in both. Not translated afterward. Written in each.
You can check our work before you call us.
Most firms ask you to trust the pitch. We would rather hand you something you can open. The three below, you can check.
Our Lab
We published the benchmark
For a month we ran three ways of building an AI against a live event, registering every prediction publicly in advance. The full dataset is released and anyone can check our work.
Read the research
Our books
We wrote the book
Adapting to AGI is published. The second, on what AI cannot fix, is in progress. The method is knowable and we give it away.
Find the book
AI Tools
We ship the tools
The software our own operation runs on, published as it proves itself outside our walls. Not slideware.
See the toolsThree ways to reach us.
A person reads every message.