We advise on, build and research the AI and technology decisions that shape how a business runs.
XONIK is a consulting, advisory, research and product firm, founded in 2006. We work directly with the executives who own AI and technology decisions, across North America, Europe, MENA and Asia-Pacific, and an engagement ends when a named person on your side runs what we built.
In brief
XONIK is a consulting, advisory, research and product firm, founded in 2006. It advises executive teams on AI and technology decisions, builds what those decisions need, researches what works and develops software of its own. For most of its history it delivered work under other firms' names. Today it works directly with the leaders who own the decisions.
On this page
Why did a firm that built for others start working under its own name?
XONIK was founded in 2006 and spent most of its history delivering work under other firms' names. The agencies and consultancies we worked for held the client relationship and sold the engagement; we turned what they sold into something that ran. That work reached household names across publishing, property, telecoms, retail and regulated gaming, and the name on it was always someone else's.
Delivering behind another firm teaches you where projects really fail, because you inherit the gap between what was promised and what has to exist on the day of launch. You meet the scope that assumed the systems agreed when they did not. You meet the pilot that had a champion and no owner, and the approval that arrived with nobody assigned to carry it out. The logos on the slides changed from sector to sector. The failures did not.
The limit of working behind someone else is that you can repair the build but not the brief. By the time a problem reached the people doing the work, the decision that caused it had been made earlier, by someone we had never met. So we changed who we sit across from. Today we work directly with the leaders who own the decisions, and the work begins where the brief is written.
A decision is finished when a named person owns it, not when the room agrees to it.
What changes when the person deciding is in the room
The first conversation is different. We ask what is at stake, what has already been tried, and who will answer for the result a year from now. When the answer to the last question is nobody, that is the first thing to fix, and it is cheaper to fix before anything is built than after.
The terms are the same for every engagement. Scope is agreed in writing before work begins: which parts of the problem are in, which are out, who has the final say, and how both sides will recognize finished. Each engagement runs under an NDA. Senior people do the work, so the practitioner who scopes your problem is the one who carries it through.
Confidentiality shapes how you can judge us. We cannot offer client names or case studies, because the agreements we work under rule them out. We publish our reasoning instead: research that states its method and its limits, field notes on problems that repeat, and free tools you can run yourself, so you can test the thinking before you speak to us.
What we believe about how this work gets done
A pilot shows that a model can do a task. It does not show that anyone will depend on it.
Dependence needs an owner, an alert that reaches a person when the system drifts, and a fallback for when the model is wrong or switched off. Those are built deliberately, and they are the work.
Technology fails at the seams between systems and between teams.
A record that reaches one system and never the other, a hand-off that happens in a meeting nobody minuted: these are ordinary, and they are where cost quietly accumulates.
Governance has to sit inside the work, or it will sit in a binder.
A policy nobody can apply on a busy afternoon is not a control. Tie each control to a system and to a person who answers for it.
Part of the shortlist is now formed by assistants and answer engines.
A business can be absent from the answer a buyer receives without ever learning that the question was asked.
A claim that cannot be checked is an opinion.
State the method, the limits and what would change the conclusion. A recommendation that cannot be wrong cannot be tested either.
How the practices, the research and our own software fit together
- 01
Advisory comes first when the decision is still open.
A senior practitioner works beside the executive who has to decide, tests the plan against the systems and the organization, and writes down what to fund, what to leave alone and who owns each choice.
- 02
Consulting carries the decision into daily use.
The integration, the process change, the pilot that has to become a system. Scope is written first, the people who wrote it do the work, and it ends with a run book and a named owner.
- 03
Research answers what a brochure cannot.
How a technology behaves under your conditions, what a market is actually doing, which supplier meets criteria that were fixed before the demonstrations. Some research is commissioned. The rest we publish.
- 04
Products come from what keeps repeating.
When the same problem arrives in engagement after engagement and nothing handles it well, we build software for it. The products are in development, and each product page says what stage it has reached.
When are we the right call, and when are we not?
We say so early when a problem is not ours to solve. The situations below mark the difference.
We are a good fit when
- A decision has been agreed and nobody has been made responsible for carrying it out.
- A pilot works and nobody can say who runs it once the project team moves on.
- Systems disagree and the people closest to them each give a different explanation.
- You want an independent read before committing money, from someone with no stake in the answer.
- A choice depends on evidence a supplier's materials cannot supply.
Look elsewhere when
- The supplier is already chosen and you want the choice justified afterward.
- You want a document more than a changed operation.
- The question is a legal, tax or audit opinion, which belongs with a qualified firm in that field.
- You want to manage a group of people rather than have the senior practitioners do the work themselves.
- No one in your business has the authority to act on what we find.
The cost of working this way lands on your calendar. We need access to the people nearest the problem, someone with authority at the table, and an account of the awkward parts of the project. A diagnosis that finds the cause inside your own organization will say so.
A pattern we are called into
The assistant that worked in the pilot and has no owner in production
A telecoms operator has piloted an assistant that drafts replies for its customer service agents. The pilot team has been moved to the next initiative. The head of customer operations is being asked when it goes live, the security lead has not reviewed what the assistant can read, and the supplier's contract covered the pilot only.
- Diagnosis. We map who would own the system, what it reads and writes, and what an agent does when it is wrong or unavailable. The head of customer operations and the security lead are both in the room.
- Scope. A written scope sets out what is in, what is out and who decides. It names the owner on the operator's side before any building is agreed.
- Build. A review step for replies the assistant is unsure of, the connection to the case system, a monitoring threshold that alerts a person, and a rule for what is logged and how long it is kept.
- Handover. A run book, a named owner in customer operations, and a date to review the first full cycle of real use.
Where it endsThe assistant is in daily use with a name beside it, or the diagnosis shows that it should not go live and the operator learns that before the money is committed.
How do you judge a firm whose clients you cannot call?
A reference call is not on offer, and a case study would break the agreement it described. What we can put in front of you is the method. Read a piece of our research and check whether it says how the finding was reached and where it stops. Run one of the free assessments and see whether the result tells you something you did not already know.
Then ask us, in a first call, how we would examine your problem, and listen for a mechanism rather than a list of services. If the answer is vague, say so.
What do buyers ask when they check us out?
Why did you work under other firms' names for so long?
The agencies and consultancies we worked for held the client relationship and sold the engagement; we made it run. That showed us where projects fail, but we could repair only the build, not the brief. So we now work directly with the leaders who own the decisions.
Who does the work on an engagement?
Senior people. The practitioner who scopes your problem carries it through, and scope is agreed in writing before work begins: what is in, what is out, who has the final say and how finished is recognized. Each engagement runs under an NDA.
You cannot name clients, so how do we judge you?
By the method, which we publish. Read a piece of research and check whether it says how the finding was reached and where it stops, or run a free assessment and see whether it teaches you something. Then ask how we would examine your problem, and listen for a mechanism.
When would you tell us you are not the right firm?
Early. We say so when the supplier is chosen and you want the choice justified, when the question is a legal, tax or audit opinion, or when nobody in your business has the authority to act on what we find. The reason is fit, never workload.
Leadership
Amit Govil
Founder and CEO
Amit founded XONIK in 2006 and has led its delivery since, across digital, technology and now AI work. He writes the firm's research and reviews its market briefings each quarter.
Read his researchCareers
Work with us.
Our hires are senior practitioners who stay with a result until it works.
The Careers page says what we look for and what to send, whether or not a role is listed.
See how we hireTalk to the people who would do the work.
Describe what you are weighing in a few lines. A senior practitioner replies with a suggested first conversation.