Certified, not just seated.
Vendor credentials prove their published scope. AI-native training adds applied evidence you can inspect.
Enterprise certification works when the path is legible, the standard is consistent, and a credential has an explicit, standards-aligned portability path. LockedIn keeps that rigor and changes the evidence: graph coverage, live-reviewed artifacts, reviewer provenance, and—before professional FDE qualification—at least one candidate-owned system deployed, operated, and transferred in production. Course attendance alone never counts.
Three gates. No shortcuts.
The graph and verified backing work establish eligibility and create an in-review record. The third gate is the separate human decision that can put the credential in force.
Complete the region
Every credential maps to a region of the skill graph — and to everything upstream of it, because a region can depend on work carrying another track's name. A free signed-in account can take the assessment at any time, but you cannot claim the credential without those nodes, and nodes don't unlock by watching.
Verified artifacts
The work itself is the evidence. A live model pass or a human expert pass can mark an artifact verified; the registry identifies which performed the review. An offline heuristic can give feedback but cannot verify, and an expert spot-check is not currently required before an in-review credential is issued.
Human activation
Clearing the graph gate with verified backing work creates an in-review registry record. A separate expert or administrator decision is required before that record becomes active. Formal live-defense formats activate with vetted roster coverage; the current credential record stores status, not a defense transcript or examiner identity.
A gate-qualified claim issues an in-review record. Only a role-gated human action can make it active; attendance moves neither state.
Seven published specifications.
Each authored standard defines a graph region, upstream requirements, backing work, and an assessment specification. Issuable credentials begin in review after graph coverage plus verified work and still require human activation. FDE-P remains visibly planned and runtime-blocked until typed field gates, live defense, and supervised evidence exist.
AI Native Foundations
Certified AI-capable: runs daily work on AI workflows with calibrated trust and hardened prompt systems.
- — 5 nodes: the L0–L1 foundations region
- — A verified artifact at every node, ending in a personal AI operating system
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — unscripted foundations review after vetted examiner coverage
Certified AI Native Developer
Certified to build with agents: spec-driven development, agentic TDD, and production agent systems with evals.
- — 14 nodes: L0–L1 foundations, agent anatomy through evaluation, developer track through L3
- — Reviewed capstone evidence: an agent release package with an eval suite
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — architecture, evals, and rollback after vetted examiner coverage
Certified Agentic Architect
Certified to design agentic systems: tool contracts, orchestration, evaluation, and fleet architecture.
- — 10 nodes: L0–L1 foundations, agents track through L3
- — Reviewed capstone evidence: two-agent slice + fleet architecture
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — contracts, gates, failure modes, and economics after vetted examiner coverage
Forward-Deployed AI Engineer — Professional
Qualified through an independently defended evidence chain spanning field diagnosis, enterprise integration, provider-portable AI systems, secure production operation, one supervised client mission, and owner-accepted transfer.
FDE-P remains planned. The runtime does not yet join all ten FDE evidence gates, an independent live defense, and a supervised field mission into one activation decision, so graph completion cannot issue or activate it.
- — 23 nodes: L0–L1 foundations, the required engineering, agent, and governed-delivery spine, plus the complete FDE route through L4
- — All ten FDE-S/1.0 evidence gates, including a supervised client mission, owner-accepted transfer, and integrated qualification defense
- — Published standard — graph coverage is only a prerequisite; offline heuristics and module completion cannot qualify an engineer
- — Planned runtime qualification — conjunctive mission, security, incident, supervised-field, transfer, and oral-defense gates after calibrated examiner coverage
- — Qualification requires a named field supervisor, client-designated owner, and independent human panel; no single score can average away a critical failure
Certified Governed AI Delivery Engineer
Certified to stand up governed agentic engineering in a regulated enterprise: risk-tiered controls, an independent verifier layer, an operating control plane, and an evidence chain that survives audit.
- — 16 nodes: L0–L1 foundations, the AI native dev environment, agent anatomy through evaluation, and the SDLC track through L4
- — Reviewed capstone evidence: a governed-deployment authorization package and evidence chain
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — adversarial authorization review after independent examiner coverage
Certified Healthcare AI Operator
Certified to operate AI in regulated care environments: PHI guardrails, staged validation, clinician-in-the-loop workflows.
- — 10 nodes: L0–L1 foundations, healthcare track through L4
- — Reviewed capstone evidence: clinical operations workflow + staged validation plan
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — guardrail and validation review after vetted healthcare examiner coverage
Certified AI Native Executive
Certified to lead transformation: operating models, governance that ships, honest AI economics.
- — 11 nodes: L0–L1 foundations, executive track through L4
- — Reviewed capstone evidence: board-ready transformation plan
- — Running gate — live-reviewed artifact coverage across the required graph
- — Planned defense — board-style review after vetted executive examiner coverage
Every issued record has a public status.
An issued registry ID resolves to its current status, required graph region, backing artifacts, and review provenance. The record distinguishes a live model review from a human expert review.
Schema is strict and errors are enumerated. TAG_LOCKED carries no recovery instruction — state whether to retry or escalate.
The failure reproduces from the trace alone: both rejected calls, the false success, and the arguments that caused them.
The enum fix addresses the observed guess directly. The retry-policy change is asserted but not shown in v2.
Idempotency is noted for the tag write only. Side effects of the lookup path are undeclared — declare them or mark it pure.
Mapped for portability. Issuance stays off until the evidence is real.
The six currently issuable definitions in this frozen v1 projection map into stable achievement, criteria, and graph-alignment records modeled on 1EdTech Open Badges 3.0. Planned FDE-P is excluded until a versioned portability v2 contract. These are not signed credentials, contain no recipient, and carry no claim of conformance or certification.
- ACHIEVEMENT MAPS
- 06
- MAPPED GATES
- 01
- INACTIVE GATES
- 07
- HOLDER RECORDS
- 00
Six v1 definitions, each bound to its real graph gate.
- AINO-F5 ALIGNED NODES
AI Native Foundations
Certified AI-capable: runs daily work on AI workflows with calibrated trust and hardened prompt systems.
MAPPING · f2e908cb308d
- AINO-D14 ALIGNED NODES
Certified AI Native Developer
Certified to build with agents: spec-driven development, agentic TDD, and production agent systems with evals.
MAPPING · 4a2f78d86703
- AINO-A10 ALIGNED NODES
Certified Agentic Architect
Certified to design agentic systems: tool contracts, orchestration, evaluation, and fleet architecture.
MAPPING · 0ec13d0a414d
- AINO-G16 ALIGNED NODES
Certified Governed AI Delivery Engineer
Certified to stand up governed agentic engineering in a regulated enterprise: risk-tiered controls, an independent verifier layer, an operating control plane, and an evidence chain that survives audit.
MAPPING · 9e710fcbe9c7
- AINO-H10 ALIGNED NODES
Certified Healthcare AI Operator
Certified to operate AI in regulated care environments: PHI guardrails, staged validation, clinician-in-the-loop workflows.
MAPPING · 4ca0b90bab52
- AINO-X11 ALIGNED NODES
Certified AI Native Executive
Certified to lead transformation: operating models, governance that ships, honest AI economics.
MAPPING · 4ee7f4e84ae8
Seven gates still stand between a mapping and a portable credential.
- 01mapped
achievement mapping
Versioned Achievement definitions, criteria, and graph alignments are projected from the authored credential catalog.
- 02inactive
issuer identity
Establish a durable issuer Profile identifier, creator authority, public verification method, and key-rotation policy.
- 03inactive
proof key management
Select and independently review a supported cryptographic proof format, custody boundary, signing service, recovery, and rotation process.
- 04inactive
credential status
Implement a durable credential-status and revocation method, public status resolution, due process, suspension, and renewal behavior.
- 05inactive
recipient privacy
Choose a privacy-preserving recipient identifier and consented evidence-disclosure policy with learner-controlled delivery and deletion boundaries.
- 06inactive
real evidence
Project only a real active Store credential backed by its exact graph gate, verified artifacts, review provenance, and current term; never use specimens.
- 07inactive
jsonld validation
Validate the complete signed credential against the official schema, JSON-LD safe mode, proof, time, and status checks.
- 08inactive
issuer conformance
Complete the applicable 1EdTech issuer tests and organizational approval before using conformant, certified, or compatible language.
This projection maps authored achievement definitions only. It is not an OpenBadgeCredential, is not signed, contains no recipient, and makes no claim of 1EdTech conformance or certification. Issuance remains disabled until every inactive gate is independently verified against a real, consented, non-specimen credential path.
Keep the rigor. Upgrade the evidence.
A clear pathway and portable credential matter. For AI work, the standard must also prove that the holder can design, evaluate, operate, and explain a real system.
Enterprise certification proved that legible paths and portable credentials work at scale. We keep that structure and change the evidence. Official provider programs can remain additive to this operating standard.
Read a hire like a diff.
A registry entry shows the required graph region, linked artifacts, review provenance, and whether the issued record is active, in review, or expired. It does not currently store a defense transcript or activating examiner.
Hire against evidence
A registry entry lists the graph nodes cleared, the artifacts submitted, and the review record behind them. You read the work, not a resume line.
Assess your team
Individual assessment and reviewed-work records are stored per person. An org coverage view begins only after real accounts are provisioned to that organization; the platform does not currently emit a cohort assessment aggregate.
Certify your own
A private cohort can be scoped against the same authored graph gates and public registry. No private cohort has run, and every active credential still requires a separate human status decision.