Yes. A COO without a technical team does not need to hire an AI team first. You need one workflow worth fixing, one person who owns it, one person who approves the output and a partner who builds inside the tools you already run, such as Microsoft 365 or SAP. What you must keep is the decisions, not the code.
#Key takeaways
- Implementing AI is mostly a data and process problem; the model is the easy part.
- A proposal that assumes a dedicated internal AI team is the wrong fit for most mid-sized companies.
- AI can be built into Microsoft 365, SAP and the systems you already pay for, without a new platform.
- Permissions need checking before any AI reads your files, so it cannot surface HR or finance documents to the wrong people.
- The skills your team needs are reviewing and approving AI output, not building models.
#What does a COO need instead of a technical team?
Four things that operations leaders already manage. A workflow that repeats often and can be explained on one page. An owner who knows its exceptions. An approver who signs off what the AI produces. And a measure the department already tracks, such as backlog, cycle steps or rework. Stennir's guide on how to pick the first process helps with the first one.
Start simple. Anthropic, december 2024 advises building the simplest working system and adding complexity only when it is needed. A single well-defined workflow with a person approving is easier to run without engineers than a set of autonomous agents.
#Why do AI proposals assume we have an AI team?
Many proposals are written for large companies: a data platform first, an internal AI team to maintain it, then use cases. For a company of a few hundred people that plan is a poor fit, and the honest response is to say so. A red flag is any firm that jumps to a multi-agent system or a custom platform before understanding your operations.
Ask instead who runs the workflow after go-live, and what your people do day to day. If the answer requires engineers you do not have, the proposal is not designed for you.
#Can AI be built into Microsoft 365 or SAP we already have?
Yes, and it usually should be. The AI reads from and writes to the places your team already works: Outlook, Teams, SharePoint, Excel and your SAP or other ERP. People keep their tools, and the output appears where they would look for it, waiting for approval.
Check permissions first. Many companies choose Copilot because company data stays in their Microsoft 365 tenant, then worry that loose sharing settings could expose HR files. The same applies to any AI that reads your documents. A permissions review comes before the build, not after an incident.
We are
Stennir builds one workflow into the systems you already run, keeps a person on your side approving, and hands you the accounts, documentation and off switch.
We aren't
Stennir is not a firm that asks you to hire an AI team or buy a new platform before anything useful runs.
#What does my team need to learn?
How to review and approve AI output, how to spot when it is wrong, and how to report a case that needs a new rule. Most staff already use AI: the Microsoft and LinkedIn Work Trend Index, may 2024 found 75% of knowledge workers use it at work, but only 39% of users had received training from their company. Role-specific training on the real workflow closes that gap, which is what AI training at Stennir does.
#What should a COO without a technical team do next?
If you know which workflow to fix, go to AI implementation. If not, AI consulting starts with an AI Opportunity Audit that ranks five workflows and recommends one build; the report is yours whether or not you build with Stennir. There is more at AI for operations. You can take the free readiness assessment or book a 30-minute call.
#Common questions
Do we need to hire anyone before we start?
No. You need an owner for the workflow, an approver for the output and someone in IT who can grant access to the systems involved. Stennir does the building and the running after go-live. If you later want an internal engineer, the code and documentation are already yours, so they can take over.
What stays with us?
The decisions, the approvals, the accounts and the data. The cloud and model accounts are in your name, the code sits in your repository, and you hold the off switch. Your team decides which process changes and signs off the output. Stennir builds and monitors inside that.
Does it work in Arabic as well as English?
Yes. The workflows Stennir builds handle Arabic and English inputs and outputs, including documents and messages that mix both. Other languages are available on request. Testing uses your own real cases in both languages, so you see how it performs on your material rather than on a demo.
What could go wrong without technical staff?
The main risk is not noticing when quality drifts. That is why every action is logged, a named person approves output, and Stennir monitors results after go-live and reports them in plain terms. The second risk is loose file permissions, which is checked before the AI reads anything.
