Skip to main content

Workflow Guide by Vertical

Relay supports ten core verticals. Each section highlights the goal, recommended workflow, and where to learn more in the public docs.

Vertical Overview​

VerticalPrimary GoalKey Capabilities
AI / ML ModelsShip reliable model orchestrationProvider abstraction, session-aware auth, performance benchmarks
MicroservicesAccelerate service deliveryRUFID routing, intelligent cache, observability suite
ServerlessRun event-driven workloads anywhereExecution engine, sandbox isolation, queue processor patterns
Smart ContractsBridge on-chain/off-chain logicSigning pipeline, webhook orchestration, deterministic RUFIDs
Edge ComputingProcess close to devices/usersEdge-ready execution adapters, cache warming, monitoring runbooks
ComplianceAutomate regulatory workflowsRetention engine, audit logging, quality gates
API ManagementGovern APIs end-to-endAPI key manager, gateway adapters, documentation automation
Code SharingCurate reusable functionsRUFID registry, session store, documentation tooling
Function CommerceMonetise function librariesMarketplace persistence, validation workflow, billing integrations
Performance IntelTrack latency & costBenchmark harness, dashboards, quality gate runner

AI / ML Models​

Goal: Orchestrate multiple providers with predictable latency and safe token management.

Workflow:

  1. Configure provider adapters and auth (see Authentication).
  2. Register inference logic via RUFIDs so history and ownership stay intact.
  3. Run the benchmark harness to capture cold vs. hot path metrics (scripts/performance/benchmark_harness.py).

Microservices​

Goal: Deliver modular services quickly without losing visibility.

Workflow:

  1. Wire service endpoints through the execution engine.
  2. Cache hot paths with the intelligent cache defaults (Intelligent Cache).
  3. Monitor service-level dashboards and alerts configured via Datadog.

Serverless​

Goal: Build event-driven workloads that scale by default.

Workflow:

  1. Define functions using the RUFID lifecycle and registry tooling.
  2. Schedule background work via queue/worker patterns in the execution guide.
  3. Validate deployments using the quality gates (dev_process/validation/quality_gate_runner.py).

Smart Contracts​

Goal: Connect blockchain events to off-chain automation.

Workflow:

  1. Secure signing workflows with the cosign-enabled pipeline (src/security/signing.py).
  2. Consume on-chain events through webhooks and runbooks.
  3. Track contract integrations with automatically generated documentation bundles.

Edge Computing​

Goal: Execute close to devices while keeping central governance.

Workflow:

  1. Deploy edge-compatible runtimes via the execution engine adapters.
  2. Enable cache warming tuned for edge latency.
  3. Monitor edge latency using the performance dashboards.

Compliance​

Goal: Automate regulatory controls without ad-hoc scripts.

Workflow:

  1. Apply the retention engine workflows (docs/database_utils.md).
  2. Run documentation + quality gates to generate audit artefacts (Configuration Hardening).
  3. Use monitoring runbooks for incident response and reporting.

API Management​

Goal: Securely expose and govern APIs.

Workflow:

  1. Manage API keys and JWT sessions (Authentication, JWT Sessions).
  2. Integrate gateway adapters or Istio control plane configuration (docs/gateway/ISTIO_CONTROL_PLANE.md).
  3. Publish endpoint docs using the documentation automation flows.

Code Sharing​

Goal: Curate a catalog of reusable functions across teams.

Workflow:

  1. Store functions in the RUFID registry with metadata and tags.
  2. Auto-generate documentation through doc tooling and quality gates.
  3. Share modules via access-controlled API keys.

Function Commerce​

Goal: Publish, validate, and monetise functions for partners.

Workflow:

  1. Manage listings through the marketplace service.
  2. Run validation workflow to enforce quality/security pre-publish.
  3. Track usage with billing integrations and entitlement dashboards.

Performance Intel​

Goal: Continuously improve latency, cost, and reliability.

Workflow:

  1. Schedule nightly benchmarks (scripts/performance/benchmark_harness.py).
  2. Review performance reports (see performance docs and dashboards).
  3. Feed insights into the quality gate runner to block regressions.

Common Next Steps​

  1. Run the smoke suite — python3 scripts/testing/run_smoke_tests.py
  2. Benchmark performance — python3 scripts/performance/benchmark_harness.py --samples 200
  3. Review quality gates and audit guidance — see docs/validation.md
  4. Select deployment mode — work with Relay SaaS for managed/hybrid or follow the self-hosted guide (Self-hosted)