.svg)
Built to do sourcing work,
not just talk about it
Most AI in procurement surfaces information and stops there. Kai drafts event details, orchestrates specialized agents, and actually executes.
Every sourcing task is completed faster
Decisions backed by real-time analysis
Machine-learning in action
Five ways Kai makes every sourcing event sharper
Best-practice sourcing, triggered by a single request
Describe what you need to source and Kai identifies the right agent for the job — then hands off to it to run the full event end to end. Intake, event creation, supplier selection, award. No manual steps, no configuration required.
RFI questions generated from your prompt
Describe your category or paste your template, and Kai generates the full list of RFI questions — including settings for when suppliers can respond, whether scoring is enabled, and whether bids are gated on RFI completion.
Bid windows set based on what the event needs
Kai uses machine learning to sets the bid duration and auto-close rules according to the details of each request — category, region, urgency, implied deadlines. Your event is live for the optimal duration: driving maximum compression and closing as early as possible.
One prompt to get anywhere in Keelvar
Ask Kai to find an event, pull up a rate, or launch an agent — and it takes you straight there. No menus to navigate, no training required. The right place in Keelvar, one prompt away.
Optimization scenarios built from plain language
Describe the scenario you want to model and Kai turns it into an optimisation rule — no formulas, no configuration, no specialist knowledge needed. Your team explores more options, in less time, with full confidence in what's being evaluated.
Numbers speak louder than words
time to value
to create & launch a request
lines of code
AI agent sitting across your sourcing stack
Heading
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla Nunc ut sem vitae risus tristique posuere.
What they say about us
“You don't even have to train the teams to use it. You now have [an agent] do all the work of building and creating the events for you.“
“The autonomous sourcing for unstructured data that Kai offers is really compelling. Seeing how that works in a category-agnostic scenario is exactly what we need as we move forward.“
Learn more about our solutions
Get in touch
FAQs
Straight answers to common questions about Keelvar.
Kai is Keelvar's AI Orchestrator. It sits at the centre of the Keelvar platform and coordinates a network of specialised AI agents to execute sourcing tasks end-to-end. Kai accepts natural language input — a buyer describes what they need to source — and routes that request to the right workflow, builds the event, engages suppliers, and manages the process through to award. It is the part of Keelvar that makes autonomous sourcing accessible to every member of a procurement team, not just technical specialists.
A user types or speaks a sourcing request in plain language — for example, 'I need to run a spot quote for 200 units of component X with our approved supplier list by end of week.' Kai interprets that request, determines the appropriate event type and structure, generates a bid sheet and cost formula, selects the relevant suppliers, and launches the event. No manual event-building is required. The user reviews and approves before anything goes to suppliers.
A chatbot responds to questions. Kai takes actions. When a user makes a sourcing request, Kai does not just advise on next steps — it executes them. It builds events, moves data, coordinates suppliers, and delivers results. Kai orchestrates a network of specialised agents, each trained on different sourcing tasks, so complex multi-step workflows can be completed autonomously. It is an AI that works, not one that talks.
No. Kai is ready to use from day one. It is built on more than a decade of Keelvar's sourcing expertise, so it already understands procurement workflows, event structures, and supplier engagement patterns. Teams do not need to train Kai or build custom models. They describe what they need in plain language and Kai handles the rest. This is what makes user adoption fast — there is no new system to learn in the traditional sense.
Kai continues to add new agent capabilities as Keelvar develops them, expanding the range of tasks it can execute. It also learns from the events run within your organisation — building familiarity with your supplier base, your category templates, and your preferred event structures — so recommendations and configurations improve with use. New capabilities are released on a rolling basis; customers do not need to re-implement to benefit from them.
Yes. Kai coordinates multiple AI agents in parallel, so it can manage sourcing events across different categories, geographies, and supplier bases at the same time. For large global procurement teams running hundreds or thousands of events per year, Kai makes that scale operationally possible without proportional headcount growth.