Batch Jobs
Run batch jobs, cron jobs, and event-driven workloads on demand. Per-second billing, custom containers, and full observability.
At a Glance
Job Types
Every workload pattern covered
Cron Jobs
Scheduled on cron expressions. Timezone aware. Retry policies. Missed runs catch-up option.
On-Demand
Trigger via API, CLI, or webhook. Pass payload as env vars. Immediate or delayed execution.
Event-Driven
Trigger on storage events, webhooks, DB changes, or custom events. Payload injected as environment variables.
Features
Built for production batch workloads
Custom Containers
Any Docker image. GPU support. Private registry. Any language.
Parallelism
Run N replicas. Auto-scale on queue depth. Resource limits per job.
Artifacts
Output uploaded to object storage. Signed URLs. Retention policies.
Secrets
Inject secrets as env vars or files. Vault integration. Rotation.
Retries & Timeouts
Exponential backoff. Max retries. Per-step and job-level timeouts.
Zero Egress
All traffic included. Logs, artifacts, API calls. No per-GB charges.
How It Works
From container to completed job in seconds.
Your Code, Your Infrastructure
Any container, any language, any dependencies. GPU support for ML batch jobs. Your code runs on our infrastructure — you control the code, we handle the infrastructure.
And when you need help, our engineers are a Slack message away.
Run Your Workloads on Demand
Per-second billing. Custom containers. Full observability. Zero egress.