Use AI for a job
you can name
Bring me the task that keeps repeating, the notes nobody can find or the process that lives in one person's head.
I will map how it works now, decide whether AI belongs in it, and build the smallest useful version with you. Sometimes that is a shared prompt. Sometimes it is a simple tool. Sometimes the honest answer is that AI will not help.
The tool comes second
A useful AI workflow has a clear input, a known output, a human check and somebody responsible for it. If those parts are missing, another prompt will not fix the process.
The answer might be a shared way to turn meeting notes into actions, scan a set of funding documents, draft a first pass or find information across a project. It should solve one real problem before it grows.
The aim is not fewer people. It is less repeated admin and more time for the work only your team can do.
Everyone has a different prompt
The output changes every time, nobody knows what is safe to paste in, and the useful experiment disappears when one person leaves.
One process the team understands
The inputs, checks and owner are written down. People know when to use it, when not to, and who signs off the result.
Small teams cannot afford AI busywork
In a grassroots organisation, the same person may be delivering the project, writing the report, answering the inbox and planning the next bid. A small piece of repeated work matters when it happens every week.
That is where AI can earn its place. We choose one low-risk task, make the result checkable, and see whether it really saves time before doing anything larger.
real task before a long list of imagined use cases
first version that is cheap to test and easy to change
review before an output becomes a decision or leaves the team
What we would do in a session
We use your real work. By the end, you should know which task to tackle, what the process needs and what not to automate.
Find the repeated work
We map the task as it really happens, including the awkward handovers and missing information.
Choose one sensible use
We weigh the time saved against the risk, cost and effort. Not every task should involve AI.
Write the shared method
We create the prompt or template, the review steps and a short guide the rest of the team can follow.
Scope a tool only if needed
If a chatbot is not enough, I sketch the smallest useful tool and what it would take to maintain.
Put the output somewhere useful
A workflow is no help if the result vanishes into another chat. We decide where it lives and who uses it next.
Test whether it worked
We name the owner, the guardrails and a simple measure. If it does not save time or improve the work, we stop.
How AI fits alongside SEO and operations
AI does not replace clear messaging, good processes, or strong SEO fundamentals. It makes them easier to maintain when used properly.
If your source material is messy, AI will magnify that mess. Clean pages, clear documentation, and structured planning inputs lead to far better outputs. That is why workflow design and digital fundamentals belong together.
The best results come when AI supports existing expertise: summarising, organising, drafting first passes, and helping teams move faster without removing human judgment.
Who AI Automation Is For
This is for a small team with a repeated job and no spare product department. You do not need to know which model, platform or automation tool to ask for.
- Charities and voluntary sector organisations
- Community Interest Companies (CICs)
- Community groups and local campaigns
- Arts organisations and independent venues
- Social enterprises and cooperatives
- Grassroots sports clubs and youth organisations
- Teams with repeated admin and no time for tool sprawl
- Any small organisation that wants useful AI, not AI theatre
I have built and tested AI workflows in agency, research and community settings, contributed to a 200,000-prompt study of AI citations, spoken about the work at BrightonSEO, and taught it in a room with Sheffield arts organisations. I bring the evidence and the mistakes, not a reseller deck.
Research, building and teaching“Confidence-building, practical and human, which isn't an easy thing to do in this area. I've already put the tools into practice.”
Freelancer (Creative Practice) · 5/5 — AI workflows session, Sheffield 2026
What is the task you keep repeating?
Tell me what it is and how you handle it now. I will tell you whether AI is worth bringing into it.
Frequently Asked Questions
Cited helps small teams use AI and LLMs for practical work: research summaries, drafting support, repetitive admin, planning systems, prompt libraries, and lightweight internal tools. The focus is on useful workflows with clear review steps, not novelty for its own sake.
Usually not. Many useful AI systems can start with tools you already have, plus better prompts, templates, and process design. Where a lightweight custom tool would help, we scope the smallest sensible build rather than jumping to enterprise software.
Simple improvements like prompt libraries, meeting-note workflows, or planning templates can help immediately. If a custom assistant or internal tool is needed, that takes longer, but the work still starts with a focused, low-risk use case so value appears early.
Yes. Small organisations can use AI safely when the rules are clear: avoid sensitive data where needed, keep humans reviewing outputs, document approved use cases, and start with low-risk workflows. The goal is better systems and less admin drag, not replacing judgment.