An AI layer over your Microsoft 365
Your team already lives in Outlook, Teams, SharePoint and Excel. We build the automations inside them, connected through the sign-on you already use, and the audit finds which of your workflows they pay back in first.
Start from what your team already does
OpenKit builds AI into the Microsoft 365 you already run. The audit maps what your team does in Outlook, Teams, SharePoint and Excel, then costs each workflow inside those tools rather than inside new software.
At House of Hackney, 20 hours a week of finance reconciliation is now automated. The admin agent we specified for Stow Brothers, a London estate agency, was designed to work under the agency’s own Microsoft 365 access controls.
The standard audit is £10,000 excluding VAT. Implementing what it recommends is a separate decision. OpenKit is ISO 27001 and ISO 9001 certified, holds Cyber Essentials and works to UK GDPR.
Connected like a member of your team
The automations sit on your Microsoft 365 as one layer, and each one reads only what its signed-in identity is allowed to read.
That one identity is what lets a single automation cross applications. It can read the thread in Outlook, pull the matching record from your CRM, update the spreadsheet and file the paperwork in SharePoint, with a person approving the steps that matter.
The audit decides which workflows earn that treatment first, measured in the hours each would give back.
Your Microsoft 365
- Outlook
- Teams
- SharePoint
- Excel
- Your CRM, finance and case systems
The layer OpenKit builds
- Signs in through your existing Microsoft identity
- Works inside the permissions your IT team has set
- Holds the judgment calls for a person’s sign-off
- Leaves a record of every action it takes
What your team gets back
- Routine replies drafted and filed, ready to send
- Cited answers from your own document library
- The Excel reconciliation run at volume
- The re-keying between systems done for you
The audit measures each of these in your business before anything is built. Two of the engagements behind this are published in full: the Stow Brothers pilot blueprint, which ran on a Microsoft 365 mailbox identity, and the House of Hackney AI charter, agreed before the tools were bought.
You’ve probably tried Copilot
It’s a reasonable place to start: it runs inside your own tenant and inherits the permissions your files already carry.
What a licence can’t tell you is where it’s earning its keep. The audit makes that judgment workflow by workflow: where Copilot covers the work, the roadmap says so, and where a built automation pays back more, it’s costed, so you’re deciding on numbers rather than on a renewal date.
It starts with the audit
First
AI Audit
A fixed scope agreed before anything starts, run with your team in the tools they work in. You leave with four things.
- Executive summary
- Interview and findings report
- Implementation roadmap
- Costing and benefit assessment
Then, if you want it
Embedded AI Lead
Where you want us to carry the work, an engagement reserves an agreed amount of delivery time for the priorities you set.
We agree who leads your engagement and how the work is supported. Where a handover is needed, it is planned with you so the context and documentation carry forward.
FAQ
Do we have to move off Microsoft 365?
No. The automations are built into the Microsoft 365 you already run, under the identity and permissions your IT team has set. Where a workflow needs software of its own, that is a bigger piece of work than an automation, so it is scoped and agreed separately.
Does it work with our single sign-on?
Yes, that is how it is built. The automations authenticate through the Microsoft identity your team already signs in with, so access is granted and revoked in the same place your IT team manages it today. An automation’s permissions can be reviewed and scoped the same way a person’s can.
Is Copilot enough?
For a lot of individual work it is: drafting and summarising over your own files, in the applications your team is already in. It runs out of road where a workflow crosses systems at volume, a mailbox that feeds a CRM that feeds a spreadsheet, because each step needs checking and nothing keeps score. The audit draws that line for each workflow and puts the hours it would give back next to it.
Have you built inside Microsoft 365 before?
Yes. Mailbox work under Microsoft 365 identity and access controls is evidenced in delivered work, on the Stow Brothers pilot blueprint. Automations in the rest of the suite are integration targets scoped into each engagement, and the audit names which integrations are in scope before any build commits.
How do you handle data security, GDPR, and regulated industries?
OpenKit is ISO 27001 and ISO 9001 certified and works to UK GDPR by default. Every implementation encrypts your data and restricts access by role, and we document how each piece of it is processed. For regulated sectors the work is designed to operate under your sector-specific obligations on top, whether that is the SRA for legal, the FCA for financial services, or the NHS DSP Toolkit for healthcare. Where client data is too sensitive to leave your environment, the same engagement can run on a private deployment, with the models operating inside infrastructure you control.
How long does it take?
The schedule is agreed in your proposal, alongside the scope. The work runs in two parts: interviews and analysis in the tools your team already uses, then the findings report, the roadmap and the costings. We book the sessions with the people who do the work first, because their availability is what usually stretches a timeline.
What does it cost?
The standard AI Audit is £10,000 excluding VAT, quoted once the scope is agreed. Implementing what it recommends is a separate decision. Where you want us to carry that work, an Embedded AI Lead engagement is from £5,000 a month excluding VAT, against a delivery allocation agreed with you. The fees, and what each covers, are on our pricing page.
Not ready to talk? The free AI readiness check scores where you stand in about five minutes.
What could your business do with AI?
We start with a conversation, audit where AI actually pays back, and leave you with a costed plan for what to build first.
We reply within one working day.