By Tim Pröhm, VP Digital Strategy & AI, KellyOCG
When we talk about total talent, I’ve come to expect a conversation about topics like vision, governance, and technology. But at this year’s ProcureCon Total Talent event in Amsterdam, a different theme stood out: The biggest obstacle to total talent is the way work enters the system in the first place.
Senior HR, procurement, and workforce leaders share a frustration with intake. It’s an operational bottleneck that quietly weakens everything downstream. Total talent falls short of its promise when every engagement still starts with a requisition form.
Once you recognise the problem, you can fix it with the right technology and process adjustments. In the past, technology was not quite ready for the job, but that’s changed.
Thanks to advances in AI, analytics, and the delivery of intelligence, total talent decisions can move to the front of workforce engagement processes where they belong.
The problem with the requisition form as the starting point is that it forces premature decisions. The moment a hiring manager opens the form, they must enter the type of talent they need, the channel to source it from, and the constraints around it. Permanent hire? Contractor? Statement of work? The req needs answers, now.
Unfortunately, the hiring manager answers those questions before the work itself is fully understood. Once a choice is made, it locks the process into a predefined path. The system routes the request accordingly; for the sake of a contingent worker, for example, suppliers are engaged, and workflows begin, all based on assumptions that may be outdated or incomplete.
What looks like a simple intake step is a data architecture problem. And it’s happening at scale.
This is where the concept of the “intelligent front door” resonated so strongly at ProcureCon. Reframe intake as a decision point instead of an administrative step. Instead of asking hiring managers to fill out a form, take them on an AI-guided intake journey.
Rather than forcing a user to select a predefined path, the system interprets the request in real time. Through natural language interaction, it maps the work being described to approved job categories, aligns it with current rate cards, and applies compliance rules dynamically.
There’s no need to log into a vendor management system (VMS) and navigate fields that may not reflect the reality of the role. The AI structures questions naturally as the conversation unfolds.
The result is clean, structured, machine-readable data at the point of entry. That matters because everything downstream depends on the quality of that initial data. Channel selection, supplier engagement (where needed), and compliance validation are all impacted. If the input is flawed, no amount of process optimisation will fix the outcome.
Once structured data exists, the organisation can make better decisions.
This is where decision intelligence comes into play. To determine the best sourcing channel for any given piece of work, AI needs access to four critical data points:
In most organisations, that data is spread across systems that don’t communicate with each other. HR systems track employees. VMS platforms track contingent labour. Procurement tools manage suppliers. Compliance data sits somewhere else entirely.
Find a way to connect and update that data, and the impact is immediate. Organisations stop going to market for talent they already have. They identify co-employment risks before they escalate. They negotiate supplier or talent pay rates based on real benchmarks, instead of outdated snapshots.
And most importantly, they score sourcing channels objectively.
This is the direction we’ve taken with our Kelly® Helix and Sevayo technology ecosystem. Rather than provide general insights, it provides a scoring layer that evaluates each potential path based on live data, not static rules.
Our account managers use this data to guide workforce planning and strategy for key clients. Agentic AI provides data to determine action, not just provide insight. You don’t have to guess whether your req should be for permanent hire or a contractor. Based on the data, you know.
One of the shifts in conversations at ProcureCon was about the role of AI. There’s no appetite for removing human decision-making. If anything, the opposite is true. The goal is not to replace people; it’s to stop wasting their time.
In a human-empowered model, AI handles the repetitive, structured components of workforce management: intake, data validation, channel scoring, and execution within defined guardrails.
Humans set the criteria. They define the policies. They intervene when exceptions arise. Every action is logged. Every deviation is visible. Every decision is auditable. Instead of managing administrative handoffs, teams can focus on workforce planning, supplier strategy, and risk management.
The organisations getting this right aren’t just moving faster. They’re reallocating capacity toward work that actually requires human judgment.
If you’re serious about advancing your total talent strategy, start by auditing your intake process.
Walk a single requisition from beginning to end. Count the steps. Track the handoffs. Measure the time it takes before work begins. That number is your baseline. And it’s probably larger than you’d expect.
If your platform cannot score channel decisions using live, connected data today, explore a new approach with KellyOCG. Test what happens when intake is structured correctly and decisions are driven by real data. Because once the entry point is fixed, everything else starts to move.