AI at Work
How to use AI at work
You're using AI for emails and social posts. The gap between that and what AI could do for the tasks that take the most time is where most organisations get stuck.
How an engagement runs
Discovery call
30 minutes, no commitment
Scoping workshop
One structured day on your real processes
Written findings
Report, priorities and a clear order of action
Delivery
Fixed price, defined deliverables, governance built in
Handover
Documentation your team can run without us
Ready to talk?
30-minute call. No obligation.
You're already using AI, so why isn't it saving time on the work that takes longest?
DSIT research published in January 2026 found 71% of UK organisations cannot identify an AI use case relevant to their work, even when they are already using AI tools. The barrier is not knowledge. It is time. Most organisations are too busy running the existing workflow to step back and redesign it from within.
For most teams, AI has settled into a narrow routine. Emails get drafted faster. Social posts come together more quickly. Then the screen closes and everyone returns to the same manual processes: compiling weekly reports from three separate spreadsheets, turning planning consultation responses into structured officer summaries, processing thirty inbound supplier quotes into a format anyone can compare.
Those tasks stay untouched. Not because AI cannot help with them, but because working out whether it can, and what it would cost to commission a solution, takes time and expertise that most leadership teams do not have spare.
An AI at Work session does that assessment for you. If you want an indicative business case before commissioning it, the free AI Value Navigator baselines your busiest workflows in about seven minutes.
For charities, the numbers tell the same story. The Charity Digital Skills Report 2025 found 76% of charities are using AI tools, but only 2% are doing so in any way that connects to strategy or saves meaningful time. Staff are experimenting without governance, without oversight, and without any link to the organisation's most demanding work.
How does an AI at Work session work?
An AI at Work session is a structured discovery and scoping engagement priced at £1,500 + VAT. We sit with your teams and work through the tasks that consume the most time in your organisation. Each task goes through a three-pathway assessment. Every recommendation passes through a governance gate before it reaches the written report.
There is no template agenda. The session starts from your actual workflows, in your own words.
Pathway one: prompt or configured AI assistant
Where a task involves producing, drafting, reviewing, or processing text and content, the session assesses whether a well-engineered prompt or a configured AI assistant is the right approach. A configured AI assistant is a purpose-built tool set up within a platform such as ChatGPT, Claude, Microsoft Copilot, or Gemini, given specific instructions to handle a defined job consistently.
The assessment covers what the prompt or assistant would need to do, what data or context it draws on, what format the output should take, and what human review the output requires before it is acted on. The post-session report contains a scoped prompt brief and a fixed-scope quote.
In practice
A North West service business with around 20 staff brought one recurring task to a session: drafting client-facing progress reports from internal project notes. The same structure was needed every time, covering milestones reached, work outstanding, and next steps, but each report was being written from scratch. It was taking three to four hours per week across the team.
The session assessed this as a Pathway One task. A configured AI assistant scoped to the firm's report structure and approved terminology, working from the project notes each account manager already kept, would handle the drafting. The post-session report contained a prompt brief built around the firm's own template and a fixed-scope quote for setup and testing.
Pathway two: automation
Where a task involves moving data between systems, triggering actions based on set conditions, or compiling information from multiple sources on a schedule, the session assesses the workflow, the systems involved, the trigger logic, and what human review the automated output requires before being acted on.
The session scopes the automation and documents the governance requirements. The post-session report contains a workflow scope and a fixed-scope quote.
Pathway Two tends to produce the clearest time savings. Public sector teams generating officer summaries from planning consultation responses, and SME operations teams processing inbound supplier quotes into a consistent comparison format for procurement decisions, are both tasks that route through this pathway. The automation is not built during the session. The session produces the scope and the quote. The build is commissioned separately from the report.
Pathway Two tasks are delivered through the AI Automation service.
Pathway three: bespoke build
Some tasks do not fit anything available off the shelf. The session assesses what the requirement involves, which systems it needs to integrate with, and what the build approach would be. This pathway produces a development brief and a fixed-scope quote in the post-session report.
For organisations that want to see what a solution looks like before committing to a full build, we offer a Try before you build service. We take the problem from the development brief, build a working version, and let you see the solution before approving the full project. If you go ahead, the cost is deducted from the final price.
Bespoke builds are scoped and delivered through the AI Software Development service. Nothing is commissioned until the client reviews the report and decides to proceed.
Every task is assessed for data risk before a recommendation is made
Before any recommendation is made, each task is assessed for data risk. The assessment covers the data involved, the UK GDPR (UK General Data Protection Regulation) basis for processing it, the human oversight requirement at each stage, and the escalation procedure if something goes wrong. Tasks that cannot proceed safely are documented with the reasoning. That record is part of the report.
This assessment runs as a gate throughout the session, not as a sign-off appended to recommendations already made. If a task involves personal data, beneficiary records, sensitive operational information, or any data that creates regulatory or ethical risk for the organisation, that surfaces during the assessment itself.
The resulting documentation is task-specific. It names the actual data sources involved, the lawful basis for processing under UK GDPR, the human review step required before the output is used, and what the procedure is if an automated source fails or produces unexpected output. A board, the Information Commissioner's Office (ICO), or a funder looking at how the organisation approached AI adoption can see exactly what decisions were made and why.
Where the session surfaces that no organisation-wide AI policy exists, that is documented as a finding. Addressing it sits within the AI Ethics and Governance service. The task-level governance documentation in the session report is not a substitute for an AI policy. The report makes that distinction clearly, and links to what an appropriate next step looks like.
What the post-session report contains
Each session focuses on three to five tasks. That is the range that allows us to assess each one properly, document the governance considerations, and produce a fixed-scope quote you can act on. Clients who arrive with a longer list are encouraged to prioritise beforehand. The preparation question in the FAQ below covers how to do that.
After the session, we produce a written report covering every task assessed: the pathway recommendation for each, the governance documentation specific to each workflow, and fixed-scope quotes for the recommended work.
Every element is written to stand on its own. Pathway reasoning, data risk assessment, human oversight requirement, and a fixed price for the work: all in a format that goes directly to a leadership team, board, or trustees without requiring further translation.
Builds are commissioned separately, after the report, and only if you decide to proceed.
Your task, the pathway, and what the report contains
Prompt or configured AI assistant
Your task
Writing reports, emails, templates, or content
What the report contains
A scoped prompt brief and fixed-scope quote
Clearest time savings
Automation
Your task
Moving data between systems or compiling from multiple sources
What the report contains
A workflow scope and fixed-scope quote
Bespoke build
Your task
Something specific to your organisation that no existing tool handles
What the report contains
A development brief and fixed-scope quote
Is this right for us?
| Right for you if | Not right for you if |
|---|---|
| You are already using AI but only for basic tasks | You have not yet decided whether to adopt AI |
| You want specific, costed recommendations for your actual workflows | You need an organisation-wide AI strategy first |
| Your team spends hours on tasks that feel automatable | You need staff training before exploring implementation |
| You need governance documentation for any AI use you commission | You are looking for a general AI awareness session with no outputs |
| You want to know what to build before committing to a larger investment | You need a same-day working build without a scoping stage |
If a broader pattern of AI adoption opportunities surfaces during the session, the natural next step is AI Strategy. The report will flag this where it is relevant.
75% of UK organisations that have adopted AI report improved workforce productivity (DSIT, January 2026). The session identifies which specific tasks in your organisation are most likely to produce that result, and what it would cost to commission the work.
Frequently asked questions
What do we actually leave with at the end of the session?
The session produces the assessment. After the session, we produce a written report covering every task assessed: the pathway recommendation for each, the governance documentation specific to each workflow, and fixed-scope quotes for the recommended work. Nothing is built on the day. The report is the document you take to your leadership team, board, or trustees to make a commissioning decision.
Will this work with the specific tools and systems we already use?
The session assesses your actual systems. Pathway recommendations are based on what you are already running, not on a preferred technology stack. Automation designs use tools built to integrate with the platforms most UK SMEs, charities, and public sector teams run already. If a system you use creates a technical constraint, the session identifies it.
How long does a session take and how many tasks can we cover?
The session runs as a full working day, on-site or remote. The number of tasks covered depends on their complexity. A session typically produces pathway assessments and governance documentation for three to five tasks. More complex tasks, particularly those routing to bespoke build, take longer to assess than tasks that resolve to a prompt or a straightforward automation.
What is in the post-session report?
The report covers every task assessed. For each task it includes the pathway recommendation and the reasoning, the governance assessment covering data handling and human oversight, and a fixed-scope quote for the recommended work. It also documents tasks where AI was assessed and found to be unsuitable, with the reasons. Every element is written to go directly to a leadership team, board, or trustees without requiring further translation.
Is this suitable for a whole team or just senior leaders?
The session works best with the people who understand the tasks being assessed. That is usually a mix of operational leads and the senior decision-makers who will authorise what gets commissioned from the report. No AI knowledge is needed to take part. The session starts from the work itself.
Who runs the session, and do they understand our sector?
Sessions are delivered by a member of our team with relevant sector experience. We match the person to the engagement based on your sector and the tasks being assessed. The session structure is designed to work from your workflows rather than a generic template, so sector-specific understanding is developed through the session itself rather than assumed in advance.
What happens next after the session, and how do we commission the work?
After the session, we produce the report and fixed-scope quotes. You review them and decide which elements to commission. Nothing is open-ended. Every piece of work quoted in the report has a defined scope and a fixed price. You can commission all of it, part of it, or none of it. If you proceed, our technical team leads delivery.
What if a task turns out to need bespoke software, and what does that cost?
Bespoke AI software and micro-SaaS builds start from £10,000 + VAT, with the exact quote based on the scope established during the session. If you want to see what the solution looks like before committing to a full build, our Try before you build service lets you do that. We build a working version from the development brief so you can approve it with confidence. If you go ahead, the cost is deducted from the final price.
What do we need to prepare before the session?
No formal preparation is needed. The session starts with what feels most time-consuming or frustrating in the current working week, and we draw the relevant tasks out from there. If you arrive with a list of more than five tasks, it helps to have a sense of which three to five to prioritise. The people who run the most manual or repetitive processes day to day are the most useful people to have in the room.
Further reading
- Learning and Tutorials
What is one of the fundamental characteristics of AI?
Artificial intelligence's defining characteristic is adaptability through learning from data rather than following fixed, pre-programmed rules.
Read more - Industry Applications
AI on the NHS: Supercomputers, safety nets & therapy bots
Britain’s NHS faces 7.8 million people on waiting lists. Traditional solutions like hiring more doctors take years to implement.
Read more - Insights and Case Studies
The Ultimate Guide to AI for Charities
Artificial Intelligence (AI) is transforming industries, and the charity sector is no exception. Charities today face mounting challenges, including funding constraints, administrative burdens, and effective donor engagement.
Read more