I know someone who spends an entire working day every month copying numbers from one spreadsheet into another to produce a leadership report. Seven hours. Every month. The report is important, the work is not. It’s simply not a valuable use of skills.
That kind of thing is everywhere, and it's the reason the conversation about digital workforce matters. Not because AI is coming for people's jobs, but because far too many capable people are spending their time on tasks that don't need a human brain.
Let's be clear about what we mean
‘Digital workforce’ sounds like it could mean anything, so let me be specific. I'm talking about AI agents, built inside Salesforce using Agentforce, that can carry out defined tasks, like pulling data together, triaging incoming queries, routing cases, compiling reports, chasing approvals. They work around the clock, they don't get distracted, and they follow the rules you set for them.
But, and this is the bit that gets lost in the hype, they are not replacing your team. They are doing the work your team shouldn't have been doing in the first place.
There's an important distinction between automation and augmentation. Automation says: "this process doesn't need people". Augmentation says: "these people deserve better work". That's the mindset shift that makes the digital workforce genuinely exciting rather than threatening.
What this looks like in the real world
Agentforce has moved beyond the pilot phase. Salesforce reported over 18,500 deals closed by the end of 2025, with its own internal support team resolving more than 80 per cent of customer queries without human intervention. Those are big numbers, but what interests me more are the practical stories.
Take the example I mentioned, the monthly reporting spreadsheet. An Agentforce agent can pull the data, format it, run the calculations, and deliver a draft report. The human still reviews it, still makes the judgement calls, still presents it to the leadership team. But instead of spending seven hours on assembly, they spend one hour on analysis. That's not replacing a person. That's giving them six hours back.
In one of the projects I'm currently involved in, we're building agents that hand over to each other throughout a process. One agent gathers the initial information, another validates it against business rules, a third routes it to the right team. The whole chain runs in the background while the team focuses on the cases that genuinely need human expertise and empathy.
UK organisations are already seeing this in practice. Police forces have deployed agents that resolve over 80 per cent of inbound citizen queries without escalation. Health companies have seen email response volumes drop by over 90 per cent through agent-first service. These aren't theoretical benefits.
The honest conversation about readiness
Here's where I want to be straight with you, because there's a lot of marketing noise around AI right now and not enough honesty.
Agentforce is powerful, but it is only as good as what you feed it. If your Salesforce instance is full of messy data, inconsistent processes, and workarounds built on workarounds, an AI agent will inherit all of that. The old saying about computers applies to agents too: rubbish in, rubbish out.
This is actually the most important thing I tell organisations when they come to us excited about AI. Before we talk about agents, let's talk about your data. Let's talk about your processes. Let's make sure the foundations are solid, because that's what turns a promising pilot into a production-ready digital colleague.
The organisations seeing the best results are the ones that treated AI readiness as an optimisation exercise first. They cleaned up their Salesforce environment, simplified their processes, and then introduced agents into a system that was ready for them.
What I'm most excited about
The thing that genuinely excites me isn't any single use case. It's the compounding effect. When you start your workday and the admin is already handled, the reports are compiled, the routine queries are triaged, and your to-do list is the interesting stuff, that changes how people feel about their jobs. Not just how productive they are, but how engaged and valued they feel.
I'm also particularly keen to explore Agentforce applications in cybersecurity. The speed at which agents can monitor, triage, and respond to security events is a natural fit, and I think it's an area where the digital workforce concept could make an enormous difference. Watch this space.
Getting started
If the digital workforce concept appeals but you're not sure where to begin, my advice is straightforward: look at where your team is spending time on work that doesn't need their brain. The monthly report no one enjoys compiling. The data entry that delays everything downstream. The routine queries that eat into your service team's capacity.
Those are your starting points. And they're often much simpler to address than you'd think.
The team we’ve built at Sopra Steria has real depth in this area. We've seen what works, what doesn't, and what the right starting point looks like for organisations at every stage of their Salesforce journey. Come and talk to us.