Question 1
Where the market boundary is drawn.
"The whole market" means nothing until someone says which market. The client will ask what was covered, and a tool that cannot show its boundary cannot help you answer.
AskWho decides what "the whole market" means, and can I see it?
A good answer: the boundary is defined with you, from geography, target companies, eligibility and inclusion and exclusion rules. It stays visible for the whole search, and "not found", "out of scope" and "not assessed" stay distinguishable.
ama's answerama maps and searches the entire relevant addressable talent market defined for a mandate across the configured sources, geography, eligibility rules, customer data, and search angles. Never miss a relevant candidate in the market you have defined. Data on more than 1 billion professionals.
Question 2
Reasoning per candidate.
A score you cannot explain is a score you cannot defend in front of a client who knows the field. The consultant answers for every name on the shortlist.
AskCan I see why each person is on the list, and argue with it?
A good answer: reasoning per criterion, with the evidence, the gaps and the uncertainty, and a way to overrule it that stays on record.
ama's answerThe Deep Context Engine is ama's assessment core. It reads careers in context rather than by keyword: role semantics beyond titles (a "CFO who sells" is not a CFO), inferred company context (funding stage, growth curve, team size and scope), and observable readiness signals. It is why ama finds the non-obvious candidate and can explain every recommendation. The three lenses, Attractability, Salary fit and Precise seniority, sit inside it.
Question 3
Data sources and hosting.
Your clients' people and your searches pass through the tool. Where the data comes from, where it sits and whether it trains a model are questions your data protection officer will ask anyway.
AskWhere does the data come from, and where does it sit?
A good answer: a written answer on sources, hosting, retention and model training, before the first search.
ama's answerPublic-data-only external sourcing. EU hosting. GDPR-compliant processing. Customer and candidate data are not used to train models.
Question 4
Sync with your system.
A tool that becomes a second system does not last. Every shortlist copied by hand into the CRM is time taken from clients and candidates.
AskDoes it land in my system, or do I get a second one?
A good answer: the objects, direction and refresh for your system, written down, and one identity per candidate.
ama's answerama works on top of the customer's system of record and syncs both ways with the named systems: SAP SuccessFactors, SAP HCM, Avature, Workday, Talentsoft, Talentrecruiter, Personio, Rexx, Ashby, SilkRoad and LinkedIn Recruiter. A system that is not on the list is built during implementation for accounts of sufficient size. The scope is recorded per customer. See the systems.
Question 5
Human control.
An agent that writes in your name is useful when you know where you steer. Every automated step should name the point where a person decides.
AskWhere do I decide, and what happens without me?
A good answer: the control point named beside every automated step: the briefing, the rules, the review, the first reply.
ama's answerYour team decides at every step that matters. ama does not make the hiring decision. With Autopilot, ama handles outreach; you take over when candidates respond.
The test
Run it on one live role.
Put the five questions to any tool on one live mandate, including ours. In ama's free role, you give feedback on the first ten profiles and count how many fit, in your own count.