What you need to know: Sourcing is often still slowed down by a lot of manual work spread across disconnected systems. Levelpath increases sourcing efficiency with AI Agents that find suppliers, draft RFPs, score bids, and route approvals, all in one system. The result is a team that can handle many times the number of sourcing events with the same resources.
Manual and disconnected work slows sourcing down
Procurement teams are taking on more sourcing events this year, often with leaner teams than last year. The good news is that the volume and the headcount do not have to move together. In one case documented by McKinsey, a company that put agentic AI to work on its sourcing made its procurement staff 20 to 30 percent more efficient. The opportunity to improve sourcing efficiency is the same for any team: give the repeatable parts of an event to software, and the hours go to the decisions that need a person.
Two things stretch out a sourcing timeline, and both come down to how the work is set up. A lot of the process of a sourcing event is still done by hand, from drafting the questionnaire through scoring responses. The work is also spread across separate systems, with the supplier list in one tool, the RFP in another, and responses arriving by email.
Delegate the manual sourcing work to AI
The most efficient sourcing event is the one that does not begin at zero. A buyer who only has a few details about what they need still gets options, because Levelpath surfaces qualified suppliers from your existing relationships, so a shortlist is ready at the beginning rather than after a week of research. Each supplier arrives with its history attached, including past spend, existing contracts, and risk signals, so the decision can be informed as well as fast.
Generating a list of questions is the easy part. The hours go into structuring them: the question types, the sections, and the conditional logic a real questionnaire needs, work that used to be done by hand. Sourcing AI Agents do that structuring now. Describe a category you have never run before, office furniture for example, and in seconds the blank fields become a working questionnaire, with short text, long text, number, and multiple-select fields, conditional logic, file uploads, and clear sections, all editable and ready to send.
As responses come in, they are automatically scored so you can quickly compare once all the bids are final.
GATX, a global railcar leasing company, has seen these results in practice. Their procurement team runs bid analyses in fifteen to twenty seconds with a single click, and can complete ten RFPs in the time one used to take. On a single bulk materials contract, AI analysis surfaced $3.5 million in savings the team could act on. Delegating that repetitive work to AI Agents is how the same team gets through many more sourcing events.
Approvals, another piece that can slow down a sourcing event, often get delayed simply because the next person did not know it was their turn. Levelpath keeps the whole event in one place: stakeholders receive automatic updates and each approval is routed directly to the person who owns it, so the work keeps moving on its own.
How much of the sourcing process can be automated?
A large portion of the sourcing process can be automated with the right tools and data. AI Agents already find suppliers, draft RFPs, score bids, and chase the responses that would otherwise sit in an inbox. What stays with people is judgment: negotiation strategy, supplier relationships, and the trade-offs that carry real risk.
Levelpath is built to let you automate as much or as little as you choose. Use pre-built AI Agents or build your own, assign them to any step in your process, and set the guardrails that fit your team. Running on the same system of record, they work from the same context a person would, so their output is accurate and data-driven. You might start with agents drafting and scoring while approvals stay in human hands, then widen what they run as your confidence grows.
The room to add more automation keeps growing. McKinsey describes agentic systems that already prepare tenders, prequalify suppliers, and analyze competing bids, which is the core of a sourcing event. Sourcing is where that shift lands first, because so much of it is structured and rules-based, and teams that hand agents the mechanical work today are the ones ready to run sourcing more autonomously tomorrow.
.jpg)
