At some point this year, someone above you or beside you — a CEO, a board member, a procurement lead, a works council — has asked some version of: “This AI thing scoring our deals. How does it work? Can we trust it? Is it safe?”
For most tools on the market, the honest answer is “we don’t really know, the vendor says it’s smart.” We think that answer is disqualifying — especially in Europe, where the bar for deploying AI at work is rising fast, and rightly so.
So here is exactly where we stand, in the same plain language we use in the product.
Advisory by design
AI in SalesSynq prioritizes human attention. It does not make decisions about people, deals, or money. The line printed in the product is the policy: “Advisory — helps prioritize human attention, not an automated decision.”
Concretely: a warning can suggest a next step, but a human takes it or leaves it. Nothing is auto-sent, auto-closed, or auto-judged. The system’s job is to make sure the right person looks at the right deal at the right time — the looking is still yours.
Show your work, or it didn’t happen
Supported Synq Score and Alert views expose the available evidence, activity, and timeline behind them. Coverage depends on the scoring path and receipt; an unavailable or incomplete view should not be presented as a finished score.
When the evidence is thin, we say so instead of bluffing. That’s what the Data trust label is: Verified when activity passed trust checks, Partial when some of it has lower confidence, Needs review when there isn’t enough history to lean on. A tool that is never uncertain is a tool that is sometimes lying.
Same inputs, same answer
Within the same scoring-rule version, configuration, eligible inputs, and evaluation time, deterministic scoring is intended to be reproducible. A changed rule, source record, freshness window, or configuration can legitimately change the result, and the publication receipt should make that context reviewable.
Generative AI can assist with supported extraction, summaries, drafts, and explanations. Its output is advisory and can be wrong; score availability and presentation remain subject to the configured scoring path and evidence gates.
Built for European expectations
We build with European transparency expectations as design inputs. Supported AI paths can apply common-pattern redaction before provider egress, but coverage varies by path and configuration, pattern matching can miss personal data, and it is not anonymisation. SalesSynq contractually does not use Customer Data to train models; provider handling depends on the enabled provider agreement and settings.
We will not stamp a compliance badge on a blog post. The AI disclosure and preliminary AI Act conformity page document the intended purpose, known limitations, and open legal review; neither is a certification or legal advice.
When someone asks “can we trust the AI?”, the right answer is never “trust us.” It’s “here’s the evidence, here’s the uncertainty, here’s the human in charge.” That’s the product we’re building.