Which AI tool is right for your business?

Sam Gareh (00:00)

Hello, and in today’s video we’re going to talk about Copilot, ChatGPT and Claude. It’s all about when to use them, whether they actually have a place within the business, and how we’ve seen these different AI models being utilised.

To start with, I wanted to touch on what they actually are. Claude and ChatGPT are probably the most popular AI models at the moment, and they’re brilliant for building and prototyping things quickly.

What we’re finding is that when you reach the point where you want to turn something into a business solution and move it into an agent for the organisation, that’s where you can start to use things like Copilot Studio and Azure AI Foundry.

What we’re looking at today is what each of these tools does best. It isn’t always about picking one and sticking with it. We can continue using different tools at different stages of the process.

I’m Sam, part of the Automation and AI Pre-Sales team at Valto. I talk to a lot of customers about this type of thing pretty much every day, so I wanted to dive into it because it seems to be a common topic we’re hearing about at the moment.

Let’s start by looking at where each of these models sits and the data they work with.

For many organisations that are already heavily invested in Microsoft, you’ll probably have used Copilot in some capacity already, whether that’s the free version included with Microsoft licences, the paid-for version, or even Copilot agents themselves.

Copilot is deeply embedded within Microsoft applications such as Word, Excel, Outlook and Teams. With Copilot Studio, you can also build your own custom agents.

ChatGPT is very much a standalone application, with an increasingly powerful agentic capability and a wide range of third-party integrations.

Claude is similar in that it’s a standalone application with strong integrations into other software and cloud-hosted services. It’s particularly strong for technical development work. Many people are already familiar with Claude Code and the concept of vibe coding, which has become a popular way of building solutions.

When it comes to pricing, Copilot can range from around £15 to £25 per user per month, although it may also be included within some higher-tier Microsoft 365 licences.

ChatGPT uses a tiered pricing model with options such as ChatGPT Business. Pricing can range from lower-cost user subscriptions through to enterprise plans that are significantly higher per seat.

Claude also generally sits around the £15 to £25 per month mark. However, what we’re seeing across the market is a move towards consumption-based pricing and credit usage models, which I’ll touch on shortly.

Prototyping vs Production

What we’re finding is that Claude, in particular, is fantastic for prototyping. This means creating quick agents and concepts so you can determine whether an idea delivers value to the business.

At Valto, we’re a Claude partner and we use it extensively for prototyping.

As an example, when we were doing a lot of work around tenders, I built both a tender scoring agent and a tender response agent in Claude. The goal was to see whether the concept worked and whether it provided value.

It absolutely did.

However, once we wanted to roll it out to the wider business, including the pre-sales team, sales team and account managers, we rebuilt the solution in Copilot Studio.

Why does that matter?

A prototype isn’t necessarily a business-ready solution. It may not meet all of the governance, security and compliance requirements that your organisation needs. That’s where Copilot Studio and Azure AI Foundry become important.

When we look at governance, an agent built in Copilot Studio or Azure AI Foundry automatically inherits your Microsoft 365 security controls.

This includes access permissions and security policies. For example, if I built a tender scoring agent but didn’t have access to a specific document required for the process, the agent would also be unable to access it.

This is why organisations often use SharePoint to create controlled knowledge bases and direct agents to the exact information they need.

Another advantage is that when moving from Claude into Azure AI Foundry, you can still use many of the same underlying AI models. This means you can maintain a similar user experience while benefiting from Microsoft 365 governance and security controls.

That’s why we’re seeing a growing trend where organisations prototype in tools like Claude and then move into Microsoft platforms for production deployment.

What Each Platform Does Best

Claude is particularly strong for technical build work and proving concepts. It’s excellent at helping users develop ideas, refine prompts and establish the foundations of an agent.

ChatGPT is what many people first encountered when AI became mainstream. It’s widely used for research, content generation and general questioning.

We still see people using ChatGPT to find specialist providers. For example, people might ask for a Microsoft partner in a specific region with expertise in particular areas, and ChatGPT can help surface relevant information.

Then we have Copilot Studio and Azure AI Foundry.

These are the platforms we typically associate with production-ready solutions. They allow organisations to take concepts and prototypes and apply governance, security and enterprise controls before deploying them at scale.

One of the biggest benefits is visibility and control over costs.

Within Copilot Studio and Azure AI Foundry, you can monitor usage and spending at a much more detailed level. If multiple agents are running through a shared ChatGPT or Claude subscription, it can be difficult to understand exactly where costs are being generated.

With Microsoft platforms, you can monitor usage by agent and even apply spending caps or credit limits.

This helps organisations maintain predictable costs and avoid unexpected bills caused by excessive usage.

The Practical Path Forward

What does this actually look like in practice?

We’re working with a growing number of organisations that have already built proof-of-concepts in Claude. They’re now reaching the point where they want to roll these solutions out across the business, but they recognise they don’t yet have the governance and controls they need.

That’s where we help move those solutions into Copilot Studio or Azure AI Foundry.

We work with organisations to understand the process they’re trying to automate, the required architecture, the information sources involved and the desired outputs.

Using the tender example again, organisations may already have a highly effective solution running in Claude. We can then help move that solution into Microsoft 365 and transform it into a production-ready agent that can be used across the wider business.

If that resonates with you and you’re currently using multiple AI tools but aren’t entirely sure which is best for a particular situation, those are exactly the types of conversations we’re having here at Valto.

If you’d like to learn more, use the link in the comments below to arrange a free consultation. We’ll take the time to understand what you’re trying to achieve and provide recommendations based on your requirements.

Feel free to get in touch or connect with me on LinkedIn. The details will also be in the comments.

Thanks for watching, and I’ll see you in the next video.