RFQ & Strategic Sourcing OS
An operating layer that runs sourcing events end-to-end: builds RFQs, gathers bids, and normalizes quotes into a clean award decision.
Where value leaks today.
A sourcing event is a long string of small frictions. Building the RFQ from the last one's leftovers. Finding the right suppliers to invite. Chasing quotes that arrive in five formats and three currencies. Then the worst part: normalizing apples-to-oranges bids into something comparable, by hand, in a spreadsheet that breaks the moment a supplier prices by the case instead of the unit. The analysis that should drive the decision arrives late and trusted only halfway.
Sourcing software sells the event but not the labor. It gives you a place to post the RFQ and receive responses; it does not write the requirements, qualify the supplier list, decode a quote that buries freight inside line items, or surface why bid B is actually cheaper than bid A once you account for payment terms and lead time. The buyer still does the thinking — and under deadline, the thinking gets thin.
So awards default to the incumbent, or to whoever quoted in the cleanest spreadsheet, rather than to the supplier who is genuinely best on total cost. Savings leak through events that were run but never truly analyzed. What's needed is a system that runs the whole event as governed execution — drafts the RFQ, qualifies and invites suppliers, ingests every quote regardless of format, and lays a like-for-like comparison in front of a human for the one call that matters: the award.
One governed flow — agents act, you approve what matters.
Every sourcing event runs to a clean, like-for-like comparison and an audit-ready award, with humans deciding the winner instead of wrestling the spreadsheet.
One operating layer — eight governed jobs.
Each is a governed agent inside the same system, sharing context — not eight tools you stitch together.
RFQ Builder
Drafts the request from the category spec and past events — scope, line items, terms, and questions — instead of copying the last one and hoping. Starts every event from a clean structured baseline.
Supplier Qualifier
Assembles and vets the invite list against capability, capacity, and compliance signals. Stops the field from defaulting to whoever you happened to email last time.
Quote Ingestor
Pulls in responses across formats, currencies, and units — case versus unit, freight in or out — and reads them into one structured schema. Ends the manual re-typing that breaks the comparison.
Bid Normalizer
Restates every quote on a like-for-like total-cost basis, folding in freight, payment terms, and lead time. Shows which bid actually wins once the hidden costs are surfaced.
Clarification Agent
Spots gaps and ambiguities in a quote and goes back to the supplier for the answer on its own. Closes the loop without a buyer babysitting the thread.
Award Briefer
Assembles the decision pack: ranked bids, total-cost rationale, risk notes, and a recommended award. Hands the human a decision, not a pile of PDFs.
Negotiation Prep
Surfaces leverage points — outlier line items, second-place gaps, prior pricing — for a final round before award. Turns the comparison into a position to push from.
Award Executor
Notifies winners and losers, captures the rationale for audit, and opens the PO against the chosen quote. Carries the decision into action instead of leaving it as an email.
Autonomy you can trust — because the control is built in.
The system acts on its own and every action stays legible, bounded, and reversible. You don't choose between speed and control; the control is what makes the speed safe.
Legible
See what was done, what was declined, and exactly what's waiting on you — nothing happens in a black box.
Bounded
Agents act only within the rules you set. Anything material or irreversible stops at a human gate.
Reversible
Every action is logged and undoable. A wrong turn is caught and rolled back, not discovered weeks later.
Owned
One operating system you own — not a swarm of rented agents you have to police. Built, run, accountable.
Every sourcing event runs to a clean, like-for-like comparison and an audit-ready award, with humans deciding the winner instead of wrestling the spreadsheet.
What you're actually getting.
Is this a product or a build?
A build. Kitsune forges the sourcing layer around your categories, supplier base, and ERP — not a generic e-sourcing seat you staff and configure yourself.
What stays in my control?
The award. Agents build the RFQ, run the field, and normalize every bid; a human makes the award call before any PO is opened.
How is this different from e-sourcing software?
Software hosts the event. This system runs the labor inside it — writing the RFQ, qualifying suppliers, decoding mismatched quotes, and laying a like-for-like comparison in front of you.
How does it compare bids that aren't apples-to-apples?
The normalizer restates each quote on a total-cost basis — freight, currency, units, payment terms, lead time — so you compare real landed cost, not headline price.
Does it handle the supplier back-and-forth?
Yes. It chases missing responses and goes back for clarifications on its own, then preps negotiation leverage, so the thread doesn't depend on a buyer watching it.
The same foundry, other domains.
Bring us the bottleneck.
We'll forge the operating layer around your friction — built, owned, and running.