Product
Eight parts, and the things each one refuses to do.
AuroWin is a decision layer, not a record-keeping layer. Its value is in what it will not let happen: an estimate presented as a fact, a price below the floor without an approval, a quotation sent on a deal that is not ready.
What each part does
The refusals are enforced in server-side code and in database constraints, not in interface copy.
Deal Passport
Turns any incoming RFQ — a PDF, a spreadsheet, an email — into the same structured record. Scope, quantities, stakeholders, deadlines, costs, competitors and the commercial conditions, in fixed fields.
What it will not do. It never presents an extracted value as confirmed. Each field carries an evidence status — confirmed, customer stated, internally observed, historically inferred, estimated, unknown or conflicting — and a separate confidence. The original document is kept as the evidence behind it.
Deal Interviewer
Reads the Passport, works out what is actually missing, ranks the gaps by how much they affect the decision, and asks only about those. Answers are written back to the field they were asked about.
What it will not do. It never asks a question whose answer it already has, never states or implies an answer inside a question, and never writes an answer into a field it was not asked about. An answer that turns out to concern a different subject is held for review instead of overwriting business data.
Deterministic pricing
Computes total cost, gross profit, gross margin and minimum selling price on the server from your configured policy, and recalculates on every read.
What it will not do. No figure here comes from a language model, and a figure supplied by a browser is rejected outright rather than merged. A price below your approved floor requires an explicit, recorded exception approval — the system will not quietly allow one.
Assessment and Estimated Win Score
Scores ten factors — customer need, decision process, economic buyer access, relationship, product fit, price competitiveness, supplier support, competitive position, delivery feasibility and evidence quality — showing each factor's supporting evidence, its missing evidence, and what would materially move it.
What it will not do. The Estimated Win Score is a structured internal indicator, not a statistically calibrated win probability, and the interface says so wherever it appears. It is never presented as a forecast.
Deal Win Plan
Produces two distinct strategies and an honest no-bid option. Each carries its pricing position, the supplier support it would need, the risks it accepts, and a red-team challenge written against its own reasoning.
What it will not do. It never guarantees a deal will be won, never invents a competitor price, a customer statement, a past deal or a relationship, and never approves its own pricing exception.
Ready-to-Quote gate
Checks the conditions that have to hold before a quotation is sound: customer identity, product, total cost, minimum price, margin policy, deadline, visible assumptions, completed approvals, and no unresolved critical warning.
What it will not do. It fails closed. When a condition is not met it refuses and names exactly which one, rather than passing with a caveat nobody reads.
Approval Inbox and Quote Workspace
Routes pricing exceptions and quote releases to the right role by threshold, and assembles the quotation with every figure recomputed from the Passport at the moment it is opened.
What it will not do. Nothing is transmitted. There is no email transport behind the Quote Workspace. Releasing a quotation records that a named person is cleared to send it themselves.
Outcome and learning loop
Records what actually happened after the quote: the price, the margin, who won, and why — keeping the customer's stated reason and the rep's interpretation as two separate claims. Patterns become organizational memory once a person approves them.
What it will not do. A single deal is not a pattern. The database refuses an organization-wide lesson that cites too few deals or too few accounts, refuses an approval by the person who proposed it, and never lets an AI-generated simulation become confirmed memory.
What AuroWin is not
Naming this saves everyone a wasted evaluation.
Not a CRM
It does not manage your pipeline, your activities or your forecast. It holds the deal information a quotation decision needs, and nothing else.
Not a chatbot
The interview asks structured questions against known gaps. It is not an open chat, and text inside an uploaded document is treated as data, never as an instruction.
Not an autonomous agent
It does not act on your behalf. Every outward step — a quotation, a supplier request, a customer message — waits for a named person to approve it.