Trust

What data moves, who decides, and what can you show us?

The questions a reviewer asks first. Below are the answers ama puts in writing, and the route for everything else.

Bring your review checklist Read the commitments
A data protection officer reads a printed document carefully.

The commitments, in exact wording.

Nothing on this list is paraphrased. Deployment specifics come from the current facts sheet for your account.

  • GDPR-compliant processing.
  • EU hosting.
  • Human oversight and governance designed for EU AI Act compliance.
  • Bias-mitigation controls in assessment and prioritization.
  • Public-data-only external sourcing.
  • Enterprise SSO/SAML support.
  • Customer and candidate data are not used to train models.
  • Security, retention, deletion, audit, data-flow and connector documentation through the enterprise review path.

The human decision map.

Where ama works, and where your people review, override and decide.

Stageama doesYour team decides
Brief and market definitionStructures criteria, surfaces assumptions, organizes the available market evidence.The business requirement, the target-company strategy, the off-limits rules, the definition of success.
Mapping and assessmentMaps and assesses inside the jointly defined addressable market.Validating the scope, adding relationship knowledge, correcting context ama cannot see.
PrioritizationShows evidence, strengths, gaps, uncertainty and trade-offs.Challenging the ranking and deciding where human attention goes.
EngagementPrepares contextual outreach and keeps status and context. With Autopilot, ama handles outreach.Voice, approval, channel and timing. With Autopilot, you take over when candidates respond.
RecalibrationShows how the pool changes, and the evidence behind it.Whether to change the brief, the offer or the strategy.
Handoff and outcomeKeeps evidence and context through the agreed workflow.Final selection, negotiation, the hire or placement, and the system of record. ama does not make the hiring decision.

The data boundary.

External discovery

Public data only. This is the boundary for finding people outside your own systems.

Your own data

Candidate, applicant, CRM or ATS data you provide is processed only for the workflow agreed with you, under the rules set for it.

Model training

Customer and candidate data are not used to train models.

Every second Personalberatung expects negative effects from regulation such as the AI Act.

Your reviewers are right to ask early. The answers should be ready before the first search.

Source: BDU, Bundesverband Deutscher Unternehmensberatungen.

The review route.

Each reviewer gets their answers from a written record, with a named owner for every open question.

ReviewerWhat they askWhere the answer comes from
IT and securityArchitecture, hosting, access, SSO, subprocessorsThe deployment record for your account
Data protectionData categories and sources, retention, deletion, rights requests, the data processing agreementThe commitments above in their exact wording, and the deployment record
Works councilWho is affected, who decides what, automated decisionsThe human decision map, and a written explanation of the intended use
ProcurementPrice, term, what the free role commits you toThe offer document that arrives with the free role
System ownerWhich objects move, in which direction, who owns a failureThe implementation record for your account

What the deployment record contains

  • Architecture and data flow
  • Subprocessors
  • Retention and deletion
  • Connector scope
  • SSO/SAML setup
  • Who answers which question

Your review contact

The review documents go to your DPO on request. Send your checklist, and each line gets a written answer with a named owner.

Pavel Möritz, Managing Director
support@heyama.ai

Review it on one live role.

One free role, run end to end with the ama team, with your reviewers in the loop from the first day.

Start with one free role Or send your review checklist first