Integrations

Plug Compute into the tools you already use.

We ship a small set of first-party integrations and document a clean API for the rest. Anything labelled “Phase 2” is on the roadmap but not built yet — see the changelog for milestone targets.

CI / CD

Submit Compute jobs from your existing pipelines — no migration required.

GitHub Actions

Beta

Drop-in workflow step that submits a Compute job and forwards logs back to the runner.

# .github/workflows/build.yml
- uses: tanvrit/compute-action@v0
  with:
    job: ./jobs/build.yaml
    workspace: ${{ secrets.TANVRIT_WORKSPACE }}
  env:
    TANVRIT_TOKEN: ${{ secrets.TANVRIT_TOKEN }}

GitLab CI

Beta

tanvrit-compute CLI in a container image; submits jobs from .gitlab-ci.yml.

# .gitlab-ci.yml
build:
  image: tanvrit/compute-cli:0.1
  script:
    - tanvrit-compute jobs run jobs/build.yaml

CircleCI

Phase 2

CircleCI orb wrapping the CLI; same UX as GitHub Actions.

Infrastructure as Code

Declare workspaces, agents, and webhooks alongside the rest of your infra.

Terraform

Phase 2

Provider exposes workspace, agent_pool, webhook resources.

terraform {
  required_providers {
    tanvrit = { source = "tanvrit/tanvrit", version = "~> 0.1" }
  }
}

resource "tanvrit_workspace" "prod" {
  name   = "acme-prod"
  region = "us-west"
}

Pulumi

Phase 2

Pulumi package wrapping the same provider for TS / Python / Go.

Observability

Forward agent metrics, job state transitions, and audit events to your existing stack.

OpenTelemetry exporter

Phase 2

Agent emits OTLP traces / metrics / logs to a configured collector.

# ~/.tanvrit/agent.yaml
otel:
  endpoint: https://otel.example.com:4317
  insecure: false

Datadog forwarder

Phase 2

Workspace-level forwarder mirrors metrics + audit log into Datadog.

Cloud storage

Per-job artefact storage backends — read inputs, write outputs.

Amazon S3 / Cloudflare R2

Beta

Read + write with IAM role / R2 token assumed at job start.

spec:
  artifacts:
    inputs:
      - s3://acme-inputs/datasets/v3.parquet
    outputs:
      - r2://acme-outputs/runs/${JOB_ID}/

Google Cloud Storage

Beta

Read + write via workload identity federation.

Azure Blob Storage

Phase 2

Read + write via managed identity / SAS.

Auth / SSO

Team and Enterprise tiers only.

OIDC SSO

Phase 2

Standard OIDC discovery; group claims map to workspace roles.

SAML SSO

Phase 2

SP-initiated SAML 2.0; group attribute -> role mapping.

SCIM provisioning

Phase 2

User + group lifecycle managed by your IdP.

Need an integration we do not list?

The REST API covers everything the CLI can do. If you build something useful, we will happily list it here — email partnerships@tanvrit.com.