Skip to main content
Jira progress: loading…

ZTS

ZAYAZ Technology Stack Specification

1. Purpose​

This specification defines the approved technology stack for all ZAYAZ platform development.

The objective is to provide:

  • Architectural consistency
  • Agent-safe development
  • Reduced technical debt
  • Predictable operations
  • Security standardization
  • Compliance traceability
  • Long-term maintainability

ZTS is the authoritative source for technology selection within the ZAYAZ ecosystem.


2. Core Principles​

Principle 1 — Strategic Standardization

Prefer platform-wide standards over local optimization.

Principle 2 — AI-First Development

All technologies must be suitable for agent-assisted development.

Principle 3 — Cloud Neutrality

Avoid unnecessary vendor lock-in.

Principle 4 — Governance by Design

Every technology decision must support:

  • Traceability
  • Security
  • Auditability
  • Compliance

Principle 5 — Registry Driven Architecture

Applications consume:

ZIR
ZCT
ZRR
ZAR
HNC

rather than hardcoded relationships.


3. Approved Languages​

Strategic

languages:
primary:
- TypeScript
approved:
- SQL
- GraphQL
- YAML
- JSON
specialized:
- Python
prohibited:
- PHP
- Ruby
- Perl

Rationale:

TypeScript becomes the universal platform language.

Python is permitted for:

  • AI
  • Data science
  • Analytics
  • Modeling
  • Computation Hub

4. Frontend Stack​

frontend:
framework:
technology: Next.js
language:
technology: TypeScript
ui:
technology: ShadCN
styling:
technology: Tailwind
forms:
technology: React Hook Form
visualization:
technology: Apache ECharts
state_management:
technology: Zustand
authentication:
technology: Keycloak

5. Backend Stack​

backend:
runtime:
technology: Node.js
framework:
technology: NestJS
language:
technology: TypeScript
validation:
technology: Zod
api:
primary:
- GraphQL
secondary:
- REST
eventing:
pattern:
- Event Driven

6. Database Architecture​

Transactional Store

transactional_store:
technology: PostgreSQL
purpose:
authoritative business data

Vector Store

vector_store:
technology: pgvector
purpose:
semantic retrieval
agent context
ontology search

Graph Store

graph_store:
technology: Neo4j
purpose:
dependency graph
blast radius analysis
ontology traversal
lineage analysis

Analytics Store

analytics_store:
technology: ClickHouse
purpose:
ESG analytics
KPI computation
reporting acceleration

Object Store

object_store:
technology: S3 Compatible
purpose:
reports
evidence
DAL artifacts
workflow outputs

7. Workflow Orchestration​

Strategic Standard

workflow_engine:
technology: Temporal

Temporal SHALL orchestrate:

Hecate
ZARA
MICE
TrustGate workflows
Verification workflows
Report generation
Computation Hub jobs

No custom orchestration engines should be introduced without architecture approval.


8. Event Platform​

event_platform:
technology: Kafka

Usage:

Domain Events
Workflow Events
Audit Events
Assurance Events
Agent Events

Mandatory patterns:

patterns:
- transactional_outbox
- schema_registry
- idempotent_consumers

9. Search Platform​

search:
technology: OpenSearch

Usage:

Document Search
Ruleset Search
Schema Search
Component Search
Evidence Search

10. Identity & Access​

iam:
technology: Keycloak

Capabilities:

SSO
OIDC
RBAC
ABAC
External Federation

Future:

TrustGate Integration
Verifier Federation

11. Agent Platform​

Strategic Agents

agents:
intake:
- Hecate
architecture:
- ZARA
orchestration:
- ACP
coding:
- OpenAI Codex
- Claude Code
validation:
- Workflow Test Engine
governance:
- FAGF

12. Agent Development Policy​

Agents MAY:

allowed:
- code generation
- unit tests
- integration tests
- workflow tests
- documentation

Agents SHALL NOT:

prohibited:
- approve contracts
- approve rulesets
- approve schemas
- bypass workflow tests
- deploy directly to production

13. CI/CD​

cicd:
platform:
GitHub Actions
deployment:
GitOps
infrastructure:
Terraform

Mandatory gates:

required:
- lint
- unit tests
- contract tests
- workflow tests
- security scan
- blast radius validation

14. Observability​

observability:
metrics:
Prometheus
dashboards:
Grafana
tracing:
OpenTelemetry
logging:
Loki

All services SHALL emit:

required:
- metrics
- traces
- structured_logs

15. Security Standards​

Mandatory:

security:
secrets:
Vault
encryption:
AES256
transport:
TLS 1.3
dependency_scanning:
required
sbom:
required

16. Technology Maturity Model​

Every technology SHALL be classified.

maturity:
strategic
approved
experimental
deprecated
prohibited

Example:

technologies:
PostgreSQL:
maturity: strategic
Neo4j:
maturity: strategic
MongoDB:
maturity: prohibited
Redis:
maturity: approved

17. Hecate Integration​

Hecate SHALL use ZTS when generating implementation plans.

Example:

implementation_guidance:
language:
TypeScript
database:
PostgreSQL
workflow_engine:
Temporal
eventing:
Kafka

18. ACP Integration​

Agent Control Plane SHALL validate generated code against ZTS.

Checks:

validation:
language_allowed
framework_allowed
database_allowed
workflow_engine_allowed
security_compliant

Non-compliant code SHALL be rejected.


19. Future Technologies​

Candidate technologies:

evaluation_pool:
WASM
Rust Services
Lakehouse Analytics
Agent Swarms
Digital Twins
GraphRAG

These may not be used in production until promoted.


20. Normative Statement​

ZTS is the authoritative implementation architecture specification for ZAYAZ.

All human developers, coding agents, CI/CD pipelines, Hecate, ACP, ZARA, and future automation systems SHALL treat this specification as the canonical source for approved technology selection and implementation governance.


APPENDIX A - Technology Decision Records (TDR)​

  1. TDR-001 PostgreSQL
  2. TDR-002 Neo4j
  3. TDR-003 Temporal
  4. TDR-004 Kafka
  5. TDR-005 Next.js
  6. TDR-006 NestJS

Future TDRs: 7. TDR-007 ClickHouse 8. TDR-008 OpenSearch 9. TDR-009 Keycloak 10. TDR-010 Vault 11. TDR-011 pgvector 12. TDR-012 GitHub Actions 13. TDR-013 Terraform 14. TDR-014 Prometheus 15. TDR-015 Grafana 16. TDR-016 OpenTelemetry 17. TDR-017 Codex 18. TDR-018 Claude Code 19. TDR-019 ACP 20. TDR-020 Hecate

Link: Technology Decision Records (TDR).




GitHub RepoRequest for Change (RFC)