HDP
Delivery Practitioner
Do the work under human control.
View curriculum →OFFICIAL HELMARK TRAINING · STD 3.0
Five certifications. One human-control model.
Build role-specific capability around FRAME → RUN → CHECK → OWN — from daily delivery and review to agent autonomy, organisational rollout and audit.
CERTIFICATION SYSTEM
Practise → Verify → Operate agents → Lead → Audit
The five paths cover different roles in the same Helmark operating model. They are role certifications, not a seniority ranking.
HDR
Delivery Reviewer
Verify AI-assisted results with confidence.
View curriculum →
HAS
Agentic Specialist
Use agent autonomy without losing human control.
View curriculum →
HDL
AI Delivery Lead
Implement Helmark across teams and delivery systems.
View curriculum →
HDA
Delivery Auditor
Assess whether Helmark control actually happens.
View curriculum →RECOMMENDED PATHS
Start with practice, then specialise by responsibility.
Delivery path: HDP → HDR or HAS
Leadership path: HDP → HDL → HDA
Equivalent practical Helmark knowledge may replace HDP where appropriate; the certificates represent different capabilities rather than a mandatory hierarchy.
PRACTITIONER PATH
Do the work under human control.
Apply Helmark to a real Work Unit from FRAME to OWN.
DAY 1
Build the model
- Why Helmark: R1–R5, Work Unit and Human Owner
- FRAME: Outcome, Owner, Boundary and Evidence
- RUN and Re-FRAME without approving every AI step
- NONE / ASSIST / CORE classification, including agent work
- LOW / HIGH / CRITICAL component criticality
- Classification + criticality case workshop
DAY 2
Apply it to real decisions
- CHECK: Right Thing? and Works Right?
- Proportional verification from ASSIST / LOW to CORE / CRITICAL
- Human Review, AI-supported review and independent CRITICAL criteria
- OWN: conscious human acceptance
- Agentic work, Cognitive Debt and Intervention Boundary
- Jira / GitHub / GitLab implementation, final case and assessment
CAPSTONE
Role-specific practical case
Take one realistic AI-assisted Work Unit through FRAME → RUN → CHECK → OWN and explain the acceptance decision.
ASSESSMENT
Certification evidence
Knowledge check + classification cases + verification decisions + final Work Unit case.
REVIEWER PATH
Verify AI-assisted results with confidence.
Specialise in CHECK, Evidence and real human review.
DAY 1
Build the model
- Reviewer responsibility inside the Helmark model
- CHECK as two questions: Right Thing? / Works Right?
- What qualifies as Evidence and what does not
- Human Review versus automated or AI approval
- Reviewing CORE work and ASSIST / HIGH or CRITICAL work
- Finding gaps between Outcome, implementation and tests
DAY 2
Apply it to real decisions
- Independent human criteria for CRITICAL changes
- Failure scenarios, runtime and observability as Evidence
- AI-supported review without outsourcing judgement
- Review disagreement and when work returns to RUN or FRAME
- Review records that prove real verification without bureaucracy
- Capstone review of a realistic CORE / CRITICAL change
CAPSTONE
Role-specific practical case
Review a simulated CORE / CRITICAL change, challenge its Evidence and decide whether it can move to OWN.
ASSESSMENT
Certification evidence
Evidence review + CRITICAL criterion exercise + role-specific final review case.
AGENTIC PATH
Use agent autonomy without losing human control.
Design bounded autonomy for agentic software delivery.
DAY 1
Build the model
- Why agent output is CORE in Helmark
- Human Owner before autonomous work starts
- FRAME for agents: Outcome, Boundary and Evidence
- Repository, data, environment, tool and permission boundaries
- Least-necessary scope without micromanaging every action
- Re-FRAME when an agent needs to expand scope
DAY 2
Apply it to real decisions
- Automated PRs, tests, reviews and scheduled agent work
- Multi-agent and agent-to-agent delegation scenarios
- Evidence for autonomous work
- Why automated action cannot constitute OWN
- Emergency work, exceptions and higher-risk boundaries
- Capstone: design and review an agentic delivery workflow
CAPSTONE
Role-specific practical case
Design a bounded agent workflow that can operate autonomously while preserving Evidence, review and human OWN.
ASSESSMENT
Certification evidence
Boundary design + agent-risk scenarios + final agentic workflow case.
LEADERSHIP PATH
Implement Helmark across teams and delivery systems.
Build the operating system around Helmark rather than adding a parallel process.
DAY 1
Build the model
- Helmark as a control layer inside the existing SDLC
- Organisation-wide NONE / ASSIST / CORE operating rules
- Component Criticality Map: LOW / HIGH / CRITICAL
- Verification Matrix and Evidence expectations
- Agent Boundary policy and Human Owner rules
- Workflow integration across Jira, GitHub and GitLab
DAY 2
Apply it to real decisions
- CHECK Load as the new AI-delivery bottleneck
- CHECK FIRST and control of parallel AI output
- Cognitive Debt, Knowledge Anchors and Handover
- Intervention Boundary and sustainable human control
- Lead Time, CHECK Load, Return Rate and L1 → L2 → L3
- Rollout workshop: create a Helmark adoption plan
CAPSTONE
Role-specific practical case
Create a practical rollout design for one product area, including criticality, verification, agent boundaries and an L1→L3 path.
ASSESSMENT
Certification evidence
Control-model design + rollout plan + leadership scenario review.
AUDIT PATH
Assess whether Helmark control actually happens.
Audit real control, not policy statements.
DAY 1
Build the model
- Conformance versus declared compliance
- Sampling real Work Units and tracing the Human Owner
- Auditing NONE / ASSIST / CORE and component criticality
- Checking FRAME, Boundary and Evidence artefacts
- Testing whether Human Review is real
- L1 and L2 audit scenarios
DAY 2
Apply it to real decisions
- CRITICAL independent criteria and agent-control evidence
- Auditing Cognitive Debt, Knowledge Anchors and Intervention Boundary
- Using CHECK Load as an operational control signal
- Exceptions, emergency work and compensating controls
- L3 conformance and evidence-based audit reporting
- Capstone audit of a simulated Helmark implementation
CAPSTONE
Role-specific practical case
Audit a simulated team using a sample of Work Units and produce a concise, evidence-based conformance finding.
ASSESSMENT
Certification evidence
Sample audit + evidence judgement + final conformance case.
CERTIFICATION BADGES
Five Helmark certification paths. Five distinct credentials.
Each certification path has its own visual badge. HDP and HDL retain the original badge designs; HDR, HAS and HDA use the new matching certification badges.
HDP
Helmark Delivery Practitioner.
HDR
Helmark Delivery Reviewer.
HAS
Helmark Agentic Specialist.
HDL
Helmark AI Delivery Lead.
HDA
Helmark Delivery Auditor.
TRAINING AT A GLANCE
Five certifications. One coherent capability system.
Compare the audience, focus and format of each Helmark certification path.
| Certification | Who it is for | Training focus | Format |
|---|---|---|---|
| HDP Delivery Practitioner |
Developers, QA engineers, testers, DevOps, software engineers and practitioners who create or accept AI-assisted changes. | Apply Helmark to a real Work Unit from FRAME to OWN. | 2 days ~13.5 training hours |
| HDR Delivery Reviewer |
Senior developers, reviewers, QA engineers, testers, tech leads and people approving HIGH or CRITICAL changes. | Specialise in CHECK, Evidence and real human review. | 2 days ~13.5 training hours |
| HAS Agentic Specialist |
Developers working with agents, DevOps, platform engineers, tech leads and teams operating autonomous or semi-autonomous coding agents. | Design bounded autonomy for agentic software delivery. | 2 days ~13.5 training hours |
| HDL AI Delivery Lead |
Engineering managers, tech leads, Heads of Engineering, delivery leads and people accountable for AI delivery governance. | Build the operating system around Helmark rather than adding a parallel process. | 2 days ~13.5 training hours |
| HDA Delivery Auditor |
Internal auditors, AI governance, risk, compliance, quality, engineering governance, consultants and people verifying Helmark adoption. | Audit real control, not policy statements. | 2 days ~13.5 training hours |
TEACH HELMARK
Interested in becoming an authorised Helmark trainer?
The trainer pathway is open to professionals with practical experience across software development and delivery roles.