Appendix D. Regulations
Use the regulation chooser to pick the right operating rule-set before applying the SOPs below.
| Field | Process description |
|---|---|
| documented actions. | |
| Scope | All active vacancies, candidate processes, recruiter team rhythm, agency and client reviews where applicable. |
| Roles | The recruitment lead owns the rhythm. Recruiters own candidate records and actions. Hiring managers own business decisions and feedback. HRD owns escalations and resources. Recruiting ops owns reports and QA samples. |
| Inputs | Vacancy report, SLA report, funnel with details, candidate movement, rejection report, source report, time-to-hire, recruiter report, consent report, action log. |
| Steps | 1. Update statuses before the weekly meeting. 2. Review previous actions. 3. Discuss priority blockers. 4. Assign owner and deadline. 5. Escalate unresolved business questions. 6. Conduct the monthly audit. 7. Record the improvement backlog. 8. Update SOP and onboarding when the process changes. |
| Mandatory fields | Meeting date, signals reviewed, decision, action, owner, deadline, verification metric, escalation if needed. |
| Metrics | Action closure rate, recurring bottlenecks, SLA trend, feedback latency, QA pass rate, regulation drift instances found. |
| SOP: Recruitment artefact QA loop | |
| Field | Process description |
| Purpose | Verify the quality of candidate CVs, rejection reasons, scorecards, AI outputs, and consent-sensitive actions. |
| Scope | Weekly samples from active roles, monthly cross-team audit, mandatory review of new recruiter artefacts and new AI processes. |
| Roles | Lead or designated reviewer examines samples. Recruiter corrects the artefact. Ops updates templates. HR/legal review privacy or assessment fairness risks. |
| Inputs | Candidate CVs, interview notes, scorecards, rejection reasons, AI outputs, consent status, signal reports. |
| Steps | 1. Select a sample. 2. Check source verification. 3. Separate facts from interpretations. 4. Check reason specificity. 5. Check AI checks. 6. Check consent-sensitive actions. 7. Record the finding as a correction, training note, or SOP update. |
| Mandatory fields | Artefact, issue, risk, mandatory correction, owner, deadline, impact on training or SOP. |
| Metrics | QA pass rate, unverified claims removed, vague reason share, AI checks share, consent issues count. |
| SOP: New recruiter onboarding | |||||
|---|---|---|---|---|---|
| Field | Process description | ||||
| Purpose | Give the new recruiter practical ability to run the process to standard before independent responsibility. | ||||
| Scope | First 30 days for new recruiters, internal transitions, and agency recruiters joining the delivery team. | ||||
| Roles | Lead owns onboarding. Buddy reviews artefacts. New recruiter completes exercises. Hiring manager joins the role context session. | ||||
| Inputs | Process map, stage/status/reason glossary, screening templates, scorecards, meeting agenda, analytics reports, AI policy, personal data consent rules. | ||||
| Steps | 1. Study the process map. 2. Complete the candidate card exercise. 3. Write a screening summary. 4. Join the weekly meeting. 5. Prepare a hiring manager brief. 6. Complete an analytics exercise. 7. Pass the QA sample. 8. Receive independent responsibility for a vacancy. | ||||
| Mandatory fields | Training module, exercise, reviewer, pass criteria, feedback, next development action. | ||||
| Metrics | Time to independence, QA pass rate, missing data share, hiring manager feedback on briefs, independent responsibility date. | ||||
| 13. Maturity model: from chaos to AI-assisted operations | |||||
| Dimension | Chaos | Managed process | Data-driven management | AI-assisted operations | |
| Weekly rhythm | Meetings happen only during crises | Weekly meeting with agenda | Action closure and blockers tracked | AI prepares brief, lead reviews | |
| Monthly audit | No regular review | Monthly metrics review | Improvement backlog based on trends | AI prepares hypotheses, people decide | |
| QA | Issues found by chance | Spot checks exist | QA metrics and training notes | QA of AI outputs and policy audit | |
| Onboarding | New recruiter learns randomly and without a plan | 30-day plan and templates | Artefacts verified against standards | AI prepares exercises, human reviews | |
| Hiring manager rhythm | Managers respond irregularly | Role review and SLA | Hiring manager bottlenecks visible | Meeting briefs highlight needed decisions | |
| Agency/client rhythm | Client asks for more CVs | Client review and feedback SLA | Shortlist quality and market signals tracked | AI drafts client note, human edits | |
| SOP maintenance | Rules live in memory | SOPs documented | SOP updates linked to process lessons | Change log includes AI process control | |
| Consent/data management | Old records used casually | Consent status checked | PD consent audit rhythm | AI does not bypass consent policy |
13. Maturity model: from chaos to AI-assisted
operations
Dimension Chaos Managed process Data-driven management AI-assisted operations
Weekly rhythm Meetings happen only Weekly meeting with Action closure and blockers AI prepares brief, lead
during crises agenda tracked reviewed
Monthly audit No regular review Monthly metrics Improvement backlog based AI prepares hypotheses, people
review on trends decide
QA Issues found by chance Spot checks exist QA metrics and training QA of AI outputs and policy
notes audit
Onboarding New recruiter learns 30-day plan and Artefacts verified AI prepares exercises,
randomly and without a plan templates against standards human reviews
Hiring manager Managers respond Role review and SLA Hiring manager bottlenecks Meeting briefs highlight needed
rhythm irregularly visible decisions
Agency/client rhythm Client asks for more Client review and feedback Shortlist quality and market AI drafts client note, human
CVs SLA signals tracked edits
SOP maintenance Rules live in memory SOPs documented SOP updates linked to Change log includes AI process
process lessons control
Consent/data Old records used casually Consent status checked PD consent audit rhythm AI does not bypass consent
management policy
| Maturity checklist | |
|---|---|
| Question | If no, next action |
| Does every weekly meeting end with an owner and a deadline? | Add an action log and review previous actions first |
| Do recurring monthly issues become improvement tasks? | Create an improvement backlog with P0–P3 priorities |
| Are new recruiters checked on real artefacts? | Add buddy-QA and pass criteria |
| Are rejection reasons audited? | Hold a monthly review of vague reasons |
| Are AI outputs verified before use? | Add a "human checks AI" field and spot audit |
| Are SOPs updated after process changes? | Add a change log and quarterly SOP review |
| What to take from the chapter / where to go next | |
| What you need | Where to find it |
| Terms and short definitions | Appendix B. Mini-Dictionary |
| Ready-made regulations and SLA | Appendix D. Ready-Made Regulations and Processes |
| Working templates | Appendix E. Template Library |
| End-to-end BDM and Backend cases | Appendix F. End-to-End BDM and Backend Cases |
| Full chapter AI prompts | Appendix H. AI Prompt Library, section H3. Process Diagnostics and Operational Rhythm |
of materials and artefact index
How to use the appendices
The appendices are not meant to be read cover to cover — they are for work: find the right artefact, pick a regulation, check a term, assemble a process in ATS, and verify that the decision remains evidence-based. If you are reading the book for the first time, start with the route at the beginning. If you are implementing the system, start with Appendix D.
Navigator by task
Task Where to read What you can take into work
Launch a vacancy without chaos Chapters 1–2, Appendix D Minimum funnel, roles, SLA, one-
page hiring plan
Set up the funnel and rejection reasons Chapter 3 Stages, statuses, reasons, next-step rules
Run screening Chapter 4, Appendix E Screening script, CV screening,
fact checklist
Prepare for interviews Chapters 5–6, Appendix F Interview cycle, scorecard,
anchors, evidence chain
Make a decision Chapters 8–9 Decision rule, debrief, decision log,
feedback caveats
Protect candidate experience Chapters 10–12 Communications, offer, preboarding,
loop closure
Manage recruitment Chapters 13–16 Dashboards, role economics, weekly
and monthly rhythm
Choose a ready-made regulation Appendix D 6 variants for agency, in-house, and
team size
Find a full AI prompt Appendix H Prompt library by task: process,
criteria, screening, debrief, offer,
analytics, pipeline
Artefact Index
Artefact For whom Where to adapt Minimum ATS fields
One-page hiring plan Recruiter, HM, HRD Chapter 2, Appendix E Role, outcome, criteria,
stages, SLA, owner
Funnel map Recruiter, ops, agency Chapter 3 Stage, status, reason, next
step, owner
CV screening summary Recruiter Chapter 4 Criterion, fact, risk, motiva-
tion, next step
Scorecard Interviewers, HM Chapter 6 Criterion, method, fact, score,
confidence
| Artefact | For whom | Where to adapt | Minimum ATS fields |
|---|---|---|---|
| Decision log | Recruiter, HM, HRD | Chapter 9 | Decision, reason, facts, commu- |
| Communication template | Recruiter, agency | Chapter 10, Appendix E | Stage, promised deadline, tone, permitted feedback scope |
| Offer note | Recruiter, HRD, finance | Chapter 11 | Terms, risks, approvals, owner, deadline |
| Pre-start context handover | Recruiter, HR, HM | Chapter 12 | Start date, terms, context, risks, first day |
| Weekly note | Lead, HRD, agency account | Chapter 16, Appendix D | Blocker, action, owner, deadline, metric |
| AI prompt | Recruiter, lead, HRD, HM | Appendix H | Task, inputs, constraints, human review |
| AI prompt | Recruiter, lead, HRD, HM | Appendix H | Task, inputs, constraints, human review |
| AI prompt | Recruiter, lead, HRD, HM | Appendix H | Task, inputs, constraints, human review |
| AI prompt | Recruiter, lead, HRD, HM | Appendix H | Task, inputs, constraints, human review |
Appendix B. Mini-Dictionary and
terminology policy
Terminology policy
The Russian form is used wherever it does not distort meaning. The English term is kept where it is established in HR/ATS/AI practice, is the name of a metric or function, or appears frequently in interfaces.
If a term is needed by a newcomer, a simple explanation is given nearby and the in-chapter link leads here.
We keep Why Russian explanation
ATS Category of systems and interfaces Vacancy and candidate management
system
SLA Operational standard Agreed response, decision, or feedback
deadline
Scorecard Established in assessment/ATS Assessment card by criteria
Shortlist Established in agency and in-house Shortlist of candidates worth discussing
recruitment further
Screening Established recruitment term Initial qualification
Offer Established recruitment term Job offer
Preboarding Established HR term Period from offer acceptance to first
working day
AI-Matching Function/approach name AI hypothesis of criteria match, verified
by a human
Process terms
Term What it means Example
Managed recruitment Hiring as a system of roles, stages, data, Every candidate has a stage, owner, and
SLA, decisions, and regular reviews. next step.
Hiring manager Owner of the business need and final Confirms criteria and participates in
decision for the role. debrief.
Next step Concrete action, owner, and deadline. HM provides feedback by Thursday.
Stage Major process point where a new Screening, interview, offer.
signal appears or a decision is made.
Status Candidate's state within a stage. Awaiting feedback, interview scheduled,
on hold.
Rejection reason Classified explanation of why the Not "other", but "mandatory criterion
process does not continue. not confirmed".
Data hygiene Habit of recording accurate, necessary, and Do not record impressions instead of
ethical data. facts.
Scope change Recorded change in requirements, stages, After the first interviews a new mandatory
budget, or role conditions. criterion was added.
| Term | What it means | Example |
|---|---|---|
| Regulation | A short working rule for the process: who does what, when, and where it is recorded. | Feedback SLA and escalation owner. |
| Escalation | Passing a problem to the decision owner without blaming anyone. | Feedback overdue, new deadline or pause needed. |
| Assessment terms | ||
| Term | What it means | Example |
| Criterion | A verifiable requirement linked to role tasks. | Running enterprise negotiations for BDM. |
| Mandatory criterion | A requirement without which the role cannot be performed at the required level. | Production ownership for Backend Engineer. |
| Learnable gap | A skill or context that can be closed after start date. | Payments domain for a strong B2B SaaS BDM. |
| Assessment fact | An observable basis for a conclusion. | Candidate described a situation, action, result, and context. |
| Confidence | How much the signal can be trusted. | A high score with low confidence requires verification. |
| Signal reliability | Quality of the fact's source: how closely it relates to actual work and is repeatable. | A work sample with defence is usually more reliable than a general impression. |
| Debrief | Discussion of facts after an interview, before the final decision. | Facts first, then conclusion. |
| Level | Expected work level within the company. | Middle/Senior, where levels are described through tasks and impact. |
| Decision log | Decision record with rationale, facts, and communication owner. | Why we advance, decline, or pause. |
| Communication, offer, and preboarding terms | ||
| Term | What it means | Example |
| Candidate experience | The candidate's experience throughout the process: clarity, respect, timelines, communication quality. | The candidate knows the next step and response deadline. |
| Feedback | Message to the candidate or client on the stage outcome within company policy. | Concise, based on work criteria, no personal labels. |
| Offer risk | Risk that the candidate will not accept the offer or will leave the process quickly. | Mismatch on compensation, role, timeline, or expectations. |
| Preboarding | Preparation after offer acceptance, before start date. | Documents, access, contact, first-week plan. |
| Loop closure | Completing the process after start date or rejection: statuses, pipeline, lessons, retrospective. | A strong finalist is retained only with a permitted justification. |
| Analytics and AI terms | ||
|---|---|---|
| Term | What it means | Example |
| Funnel | Sequence of stages and candidate transitions. | New → screening → interview → offer. |
| Conversion | Share of candidates who advance to the next stage. | Screening-to-interview conversion. |
| Time-to-hire | Time from launch or first contact to hire, using the chosen methodology. | Only use with a clear start point. |
| Cost of an unfilled vacancy | Management hypothesis about losses from a delayed role. | Calculate as a range, not exact ROI. |
| AI output | Draft, conclusion, or hypothesis prepared by AI. | Can help structure notes but does not make decisions. |
| Human review | Mandatory human verification of facts, privacy, tone, and permissibility of the conclusion. | Recruiter removes unverified claims from AI CV summary. |
| Data minimisation | Share and store only data needed for the hiring purpose. | Do not include personal details in prompt if not relevant to the role. |
Appendix C. Sources,
methodological notes, and
legal caveats
How to read the sources
This book is a practical handbook, not an academic review or legal advice.
The sources below serve as a methodological anchor for careful formulations: structured interview, criteria-based assessment, human review, privacy, and non-discrimination.
Methodological sources
Topic Source / direction How it is used in the book
Validity of selection methods Schmidt & Hunter, 1998; research on Supports the idea that assessment methods
validity and practical utility of selection differ in reliability and must be linked
methods to role tasks.
Structured interviews McDaniel et al., 1994; Campion et al., Supports predetermined criteria, uniform
1997; Google re:Work materials on structured questions, and comparable candidate
interviews assessment.
Assessment fairness and EEOC materials, EU Equal Treatment Supports the prohibition on sensitive
non-discrimination Directive 2000/78/EC as general context assumptions and the requirement for
work-related criteria.
Personal data GDPR as a well-known example of data Supports the principles of data minimisation,
regulation; local data policies, consent, consent, retention periods, and access
companies restriction.
AI governance NIST approaches to AI risk management Supports the rule: AI assists but does not
and human review practices make sensitive decisions without human
verification.
HarmonyATS reference materials
Topic Where to look
Personal data consent HarmonyATS documentation / personal data and
consent section
Source report HarmonyATS / reports / source report
Rejection report HarmonyATS / reports / rejection reasons
Candidate movement HarmonyATS / reports / candidate movement
Funnel with details HarmonyATS / reports / funnel with details
Time-to-hire HarmonyATS / reports / time-to-hire
SLA report HarmonyATS / reports / SLA
AI Studio and AI-Matching HarmonyATS / AI Studio and AI-Matching
Caveats
Check applicable employment, anti-discrimination, and privacy regulations in your
jurisdiction.
External feedback to the candidate depends on company policy, note quality, and the legal
context.
The economic models in this book are management hypotheses, not financial forecasts or
ROI guarantees.
AI outputs must not be used without verification of facts, privacy, tone, and permissibility
of wording.
Appendix D. Ready-Made Regulations
and Processes
Which regulation to use as a base
Select your work type and team size, then adapt SLA, roles, and ATS fields
1 person Small team Large team
Agency solo Agency small Agency large
Agency Account owner Delivery lead Account director + ops
client weekly client memo QC gate + role owner steering review
shortlist consent + feedback SLA duplicate rules QA sampling
In-house solo In-house small In-house large
In-house HR/recruiter + HM Recruiting lead HRD + recruiting ops
business 30-min weekly review scorecard + debrief governance + AI policy
team minimal funnel monthly quality review exec dashboard
How to choose a regulation
Select the row by work type and column by team size. Then adapt SLA, roles, ATS fields, and
cadence to the hiring volume, role risk, and process maturity. Do not copy the regulation mechanically — it must
survive a high-load week.
Common baseline for any regulation
Block Minimum
Purpose Manage the role, candidate, decision, and communication without
relying on one person's memory.
Roles Recruiter, hiring manager/client, interviewer,
HRD/lead, coordinator/ops where applicable.
Funnel New candidate → screening → interview with hiring manager →
assessment → decision → offer → preboarding /
closure.
SLA Response deadline for each stage and escalation owner.
Artefacts Hiring plan, CV screening summary, scorecard, decision log,
offer note, pre-start context handover.
| Block | Minimum | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ATS fields | Stage, status, owner, next action, source, consent, reason, assessment facts, SLA date. | ||||||||
| QA | Weekly check of stalled candidates, empty reasons, weak facts, AI outputs. | ||||||||
| Six compact regulations | |||||||||
| Variant | For whom | Owner | Rhythm | Mandatory artefacts | ATS minimum | QA / escalation | |||
| Agency: single client owner | Boutique agency, one recruiter | Client owner | Daily next actions, weekly client note | Client brief, shortlist definition, consent, submission note, feedback log | Client, vacancy, source, consent, owner, next action, client feedback | No feedback for 2 days, criteria change, duplicate conflict | |||
| Agency: small team | 2–7 recruiters / account managers | Recruitment lead + role owner | Twice-weekly role review, weekly client sync | Sourcer brief, quality control, shortlist scorecard, duplicate rules | Stage owner, quality control status, feedback SLA, reason taxonomy | QC before submission, 10% submission sample | |||
| Agency: large team / RPO | Recruitment teams, multiple clients | Account director + recruiting ops | Weekly team review, monthly leadership review | Management plan, SLA matrix, risk register, data glossary | Account reporting, audit fields, consent, dashboards | Escalation to leadership review, compliance QA, monthly sample | |||
| In-house: one HR/recruiter | 1 HR/recruiter + founder / hiring manager | HR/recruiter + hiring manager | Weekly 30-minute hiring review | One-page hiring plan, minimum funnel, hiring manager SLA, candidate updates | Priority, stage/status, reason, owner, next action, consent | Hiring manager feedback over 2 days, weekly stalled candidate check | |||
| In-house: small team | 2–6 recruiters | Recruitment lead + role owner | Weekly portfolio review, monthly quality review | Interview cycle, scorecard, interviewer brief, decision log | Scorecard completion, feedback SLA, source/reason, offer fields | Missing scorecard, criteria drift, offer delay | |||
| In-house: large team | HRD / recruiting ops | HRD + recruiting ops | Weekly operational review, monthly leadership review, quarterly calibration | Operating model, prioritisation, AI policy, leadership dashboard | Mandatory fields, access rules, audit trail, AI usage log | SLA thresholds, data quality, assessment fairness and privacy checks | |||
| SLA Library | |||||||||
| Situation | Minimum SLA | Escalation | |||||||
| New relevant application | 1–3 working days | If applications pile up without review, reassess capacity or role priority. | |||||||
| Screening completed | By end of next working day | If no decision, assign a clarification owner. | |||||||
| After interview | 1–3 working days or the previously promised deadline | If hiring manager is silent, involve the lead, HRD, or client owner. | |||||||
| Offer decision | As soon as possible, typically 1–5 working days | If budget, finance, or legal approvals are needed, set a separate owner and date. | |||||||
| Candidate delay | Before the promised deadline is breached | Even without news, send a brief status update. |
| Escalation Library | |
|---|---|
| Failure | What to do |
| Criteria changed after interview | Pause new decisions, update hiring plan, flag affected candidates. |
| Client/manager does not provide feedback | Show SLA breach, assign owner, agree new deadline or pause. |
| Rejection reason is empty or weak | Do not close a stage without a reason; use the taxonomy and assessment facts. |
| AI output contains unverified conclusions | Remove conclusions, keep only verifiable facts and questions for the human. |
| Candidate is stalled without a next step | Assign an owner, a promised update date, and the next step. |
