Cloud Pickle Solutions

04/03/2026

Most businesses are not struggling because of their people.
They are struggling because of their processes.

Manual work.
Disconnected systems.
Too much admin.

At Cloud Pickle Solutions, we build tailored apps on the Microsoft platform using AI and low code tools like Power Platform and Azure.

We design solutions around how your business actually runs.

• Custom applications built for your workflow
• AI added where it genuinely improves the process
• Automation that removes repetitive tasks
• Connected systems that give you clear visibility

No squeezing your business into generic software.
No overengineering.

Just practical, well built solutions that save time and reduce friction.

If you are already using Microsoft 365, you likely have more capability available than you realise.

Take a look at www.cloudpicklesolutions.com
or send me a message if you want to see what this could look like in your business.

26/02/2026

I had to get on board with it.

Everywhere I looked, AI this, AI that. Feels like if you don’t mention it, you’re behind. So I leaned into it properly over the last 6 to 12 months and asked myself a simple question.

Where does this actually make sense inside a real business?

Not in theory. Not in a demo. In the middle of someone’s working day.

What Cloud Pickle has started doing is blending low code apps with AI in a practical way. Clean user interface where you need structure and control. AI layered in where it removes effort.

So instead of people sorting emails for an hour, AI handles it.
Instead of someone reading through pages of noise, AI summarises it.
Instead of manual steps between systems, AI bridges the gap.

And the big thing I’ve noticed?

AI is very good at giving you your time back.

That’s the shift for me. It’s not about building flashy tools. It’s about finding the balance between the sci fi version of AI and what you can use today without turning your business upside down.

The future will be heavily AI driven. That’s obvious. But the companies that win won’t be the ones chasing every trend. They’ll be the ones who apply it in the right places.

That’s where I sit now.

I’ll be sharing more soon about the solutions we’ve built that are already freeing up hours across teams, quietly, in the background.

For now, here’s the slightly exaggerated photo of me apparently living in the AI future. It’s not far off.

27/11/2025

🩷✨Our New Website Is LIVE! ✨🩷

I’m so excited to finally share that our brand-new website is now up and running — and it’s everything I hoped for and more! It feels so perfectly on brand, easy to use, and truly represents what we’re all about.

A massive thank you to Cloud Pickle Solutions for creating it. Their service was incredible from start to finish — nothing was ever too much trouble, and they absolutely nailed the look, feel and functionality I wanted. Highly recommend!

And the best part?
🎄 Our Brownie Yule Logs are now available to purchase directly through the website! 🎄
Simple, quick and stress-free ordering just in time for Christmas.

www.mummybearcakes.co.uk

18/09/2025

Ever wondered how AI “remembers” context instead of just guessing?

That’s where vector databases come in.

Think of them like a super-organised filing system, but instead of sorting by keywords, they sort by meaning.

For example:

Sentences like “Barney is furry and cute” and “Barney the fluffy dog” get stored close together because they mean almost the same thing.

A sentence like “X is a chemical compound” sits much further away, because it has little to do with Barney.

This “distance” is measured in numbers, called vectors. The closer the numbers are, the more related the meaning.

Why this matters for AI
When you ask an AI a question, it doesn’t search every word in every document. Instead, it checks the vector database to quickly pull the most relevant pieces of information — the ones closest in meaning to your request.

How businesses use it

- Customer support: AI retrieves the right answer from thousands of help articles.

- Legal and compliance: AI finds the most relevant clause in contracts.

- Research: AI links related studies, even if the wording is different.

- Internal knowledge: AI searches policies and past projects by meaning, not just keywords.

Without vector databases, AI would lose track of context. With them, it can connect the dots between “furry and cute” and “fluffy dog” in the same way your brain does.

17/09/2025

Have you been wondering what AI agents are and why everyone’s talking about them?

We’ve all tried a chatbot by now. You ask it a question, and it gives you an answer. Helpful, but a bit limited.

Now imagine something smarter—something that doesn’t just answer, but actually takes action for you. That’s an AI agent.

Here’s the difference:

Ask a chatbot: “How can I get to Miami?” It’ll list out travel options—planes, trains, cars.

Ask an agent: it checks flight times, compares routes, finds the best prices, books your tickets, adds the trip to your calendar, and even sorts out your travel insurance.

Agents don’t just predict words. They can use the tools and data you give them to plan, act, and deliver results.

How it works

LLM foundation: The brain is still a language model, predicting what comes next.

Tools: You connect it to your apps, data, or systems. That’s how it takes real action.

Planning: Agents break down tasks into steps, follow them, then report back with the outcome.

Why does this matter for businesses

They can be trained on your company’s knowledge, policies, and processes.

You decide what tone they use, what they can do, and more importantly, what they should not do.

Example: Customer service. An agent could read an incoming email, pull details from your CRM, draft a personalised response in your company’s style, and schedule a follow-up.

Example: Finance. An agent could gather data from multiple systems, reconcile numbers, create a report, and send it to the right people—no manual copying or pasting.

Where this is going

Agents today are still evolving, but they already show what’s possible. In the future, they’ll handle complex workflows from start to finish, becoming trusted digital team members.

Over the coming weeks I’ll be sharing more on how agents work and how they’ll fit into business.

I’d love to hear from you: what’s one task you’d hand over to an AI agent first?

16/09/2025

So, what’s an LLM I hear you ask?

Ever used your phone and it guessed the next word you wanted to type? That’s prediction. A Large Language Model (LLM) does the same thing, but on a much bigger scale.

An LLM is basically a prediction machine. It looks at the words you’ve typed, then predicts what should come next. Instead of learning from just your texts, it’s trained on billions of examples. That’s why it can chat, write code, summarise, or even draft stories.

Here’s the breakdown:

Models are the versions of this predictor. Each one is trained differently, like updated editions of a textbook.

Parameters are the settings inside. Picture billions of tiny switches. The more there are, the more patterns the model can capture. That’s why models with more parameters often give more natural and accurate answers.

Websites like ChatGPT.com and Open WebUI are simply the interface you use. It’s the cover of the book. The LLM itself is the pages inside, where the real work happens.

LLMs don’t think or understand the way people do. They recognise patterns in language and use those patterns to predict what comes next. Their “intelligence” comes from the scale of data and the number of parameters they’ve been trained with.

And no, they don’t know everything. They work from the data they’ve seen during training, which means they can still get things wrong.

16/09/2025

Are you wondering how to host AI in a secure and future-proof way for your business?

I’m in Manchester this week meeting with Softcat to talk about their AI infrastructure as a service offering. The big question is simple: where should you run your AI and LLM workloads?

You have three main options:

IaaS: Rent hardware and run everything yourself. You keep control, but the risk is whether that hardware can keep up. AI models are growing fast, and what works today could feel outdated tomorrow.

PaaS: Use a platform to manage the infrastructure for you. Easier to scale, but you need to trust that provider with your proprietary or financial data. How confident are you that it won’t be breached?

SaaS: Subscribe to ready-made AI tools. Quick to start, low overhead, but less customisable. You trade flexibility for convenience.

Which one is “best” depends on:

Cost – renting GPUs vs paying for managed services vs subscription.

Your resources and skills – do you have people who can manage AI infrastructure in-house?

Your use case – is it sensitive financial data, internal productivity, or customer-facing services?

For example:

A bank might keep AI inside its firewall using IaaS or PaaS because of compliance.

A startup might jump on SaaS to move fast and stay lean.

A manufacturer could use PaaS to balance control and ease while experimenting with predictive maintenance.

Over the next few weeks, I’ll share my thoughts as I compare these models and try to answer the question: where should you run your AI?

I’d love to hear what you think. Which option makes sense for your business?

09/09/2025

20/08/2025

07/08/2025

Smarter workflows. Fewer bottlenecks. Sounds good, right?

We help businesses automate the bits that slow them down — from clunky approvals to messy handovers. Using Microsoft Power Platform, we build streamlined solutions that fit right into your Microsoft 365 environment.

Less chasing. More doing.

Want to see how it could work for you? check out cloudpicklesolutions.com or drop us a message.

05/08/2025

Still stuck doing the same tasks over and over? There’s a smarter way to work.

At Cloud Pickle Solutions, we help businesses like yours move faster by building custom Power Apps tailored to your exact needs.
✅ Automate repetitive work
✅ Free up your team
✅ Integrate seamlessly with Microsoft 365
✅ Say goodbye to clunky workarounds

Whether you're buried in spreadsheets or juggling paper forms, we’ll help you build digital tools that actually make sense for your workflow.

Ready to unlock your full potential?
👉 cloudpicklesolutions.com or drop us a message today.