Ship agentic workflows to any GPU backend.
plnt is the orchestration runtime for micro-agent workflows. Pick a workflow from the registry, pick a Kubernetes GPU cluster, and plnt handles the Helm deploy, canary, and rollback — as a durable Temporal saga.
Three repos, one story. An end-user product that needs narrow, reliable AI features. A public registry of workflow recipes anyone can pull. A runtime that turns any Kubernetes cluster into a place those recipes can run.
google-business
AI copilot for Google Business Profile owners. Auto-reply to reviews, generate weekly posts, triage booking inquiries, publish updates.
The end-user product. Pulls its workflows from microagents; runs them on plnt.
View repo ↗microagents
Pluggable agentic workflows. Each one ships as a spec + code, versioned and pullable over S3. Think npm, but for agent recipes.
The workflow layer. Anyone can publish; anyone can pull. Fully open.
View repo ↗plnt
Given a workflow spec + a Kubernetes backend, plnt orchestrates the run: Helm deploy, canary, smoke test, promote or rollback, all as a Temporal saga.
The infra load-bearing piece. This is the deep repo.
View repo ↗ ┌────────────────────────────────────┐
│ google-business (end-user SaaS) │ reviews · posts · bookings
└───────────────────┬────────────────┘
│ invokes
▼
┌────────────────────────────────────┐
│ microagents (workflow registry) │ pluggable recipes on S3
│ review-responder · post-generator │
│ booking-triage · trend-monitor │
└───────────────────┬────────────────┘
│ pulls spec
▼
┌────────────────────────────────────┐
│ plnt (orchestration runtime) │ ← this repo
│ WorkflowRun CRD · Temporal saga │
│ Helm install · canary · rollback │
└───────────────────┬────────────────┘
│ helm install
▼
┌────────────────────────────────────┐
│ Kubernetes GPU backends │ kind · GKE · EKS · on-prem
│ scheduler · nvidia.com/gpu │
└────────────────────────────────────┘ Six primitives. Enough to turn a workflow spec into a running service on a GPU cluster, without hand-rolled scripts.
Workflow-first
Point plnt at a workflow spec pulled from any OCI or S3 registry. Steps, tool bindings, GPU class, replicas — all declared, not scripted.
Backend-agnostic
Any Kubernetes cluster with GPU nodes. Same YAML runs on kind, GKE, EKS, on-prem. Pick the backend at deploy time.
Canary by default
Every run is a canary first. 5% traffic → smoke test → promote. Fails fast, rolls back cleanly. Compensation via Temporal, not bash scripts.
Durable saga
Every deploy is a Temporal workflow. Process restarts resume from the last completed step. Nothing hides in bash state or in-memory queues.
Multi-GPU aware
Workflows declare GPU class + count. plnt schedules across node pools, respects nvidia.com/gpu limits, streams GPU utilization back to the caller.
One CRD, one contract
A single WorkflowRun CRD is the entire deploy trigger. kubectl apply. The operator does the rest.
Three steps. No control plane to run. No deploy scripts to maintain. The Kubernetes and Temporal primitives do the work.
Fetch a workflow spec from microagents (or any S3/OCI registry). Verified by hash, pinned by version.
plnt starts a Temporal saga: validate → helm install canary → smoke → promote or rollback. Durable, resumable, inspectable.
Workflow is reachable at <name>.plnt.svc/invoke. Uniform request shape across every workflow.
One CRD. One Helm install. One curl.
apiVersion: plnt.work/v1
kind: WorkflowRun
metadata:
name: review-responder
spec:
workflow:
ref: review-responder@1.2.0
registry: s3://microagents
backend:
cluster: gpu-cluster-01
gpuClass: nvidia.com/h100
gpuCount: 2
replicas:
min: 1
max: 4
canary:
trafficPercent: 5
smokeTest:
invocations: 10
p95BudgetMs: 2500 The playground runs live. The repos are open. The docs walk through a 60-second kind demo.