Tablize runs in 4 regions — San Jose, Tokyo, Hong Kong, Frankfurt. Pick one when you create a workspace. Your database, object storage, and MQTT broker all live in that region. Backups stay in-region too. Cross-region transfers happen only with your explicit configuration.
sjc nrt hkg fra Region is fixed at workspace creation. To change region you must create a new workspace and migrate.
Five steps, about two minutes. A screenshot-rich walkthrough lives in the quickstart guide.
Open console.tablize.com and sign up or log in.
The top-right New Workspace button opens the creation modal.
The Region dropdown lists all 4 regions: sjc (San Jose), nrt (Tokyo), hkg (Hong Kong), fra (Frankfurt). Pick the one that matches your residency requirement.
⚠️ Region is immutable after creation. If you pick the wrong one, you have to create a new workspace.
Slug is the URL identifier (lowercase alphanumeric + hyphens, 3–32 chars). Display name is the human-readable label. Click Create — a dedicated PostgreSQL, S3, and MQTT broker spin up in your chosen region within about 60 seconds.
Settings → Workspace shows the region code. If it reads, say, fra (Frankfurt, Germany), every data plane component lives in that region. Backups stay in-region too.
Region is per-workspace, not per-account. Run a US workspace for your HQ team and a Frankfurt workspace for your EU subsidiary from the same account.
Once you pick a region, all of the following live in that region's infrastructure.
These happen by default in the managed configuration. Each has a documented workaround if your compliance posture rules them out.
When third-party clouds are off the table — financial, healthcare, public sector, defense — run Tablize on your own infrastructure. Same binary, deployed via Docker Compose on AWS, GCP, Azure, Alibaba Cloud, Sakura, Oracle Cloud, OVH, or on-premises. Your network boundary, your encryption keys, your audit log.
Self-host docs →Yes. Settings → Workspace shows the region code. On request we provide Fly.io machine placement evidence and DNS dig results. Enterprise DPAs include this as an annex.
Yes. Free can pick any of the 4 regions at creation, but free workspace data is ephemeral (discarded at session end). Persistent regional storage starts at Plus.
Not in-place. Create a new workspace in the target region and re-import via exports or re-connect integrations. For large workloads contact support@tablize.com — we help with the migration.
Yes — Pro plan and up. Same binary, deployed via Docker Compose on your AWS, GCP, Azure, on-prem, or sovereign cloud (e.g. Sakura, Oracle, OVH, Alibaba). See self-host docs.
Yes. Each workspace gets its own MQTT broker (EMQX) in the selected region. Devices connect over TLS (port 8883) directly to your regional endpoint.
Two paths. (1) Bring your own LLM key — for example Azure OpenAI deployed in Frankfurt or Tokyo, or Anthropic via your own account. (2) Self-host with `BYO_LLM_URL` pointing at an on-prem Llama / Qwen / Mistral. Both keep prompts and responses inside your chosen jurisdiction.
Each region maps to Fly.io facilities (nrt = Tokyo area, fra = Frankfurt, sjc = San Jose, hkg = Hong Kong). Formal location attestation, Fly.io SOC 2 reports, and subprocessor details are available via trust@tablize.com.
Deleting a workspace purges the PostgreSQL DB, S3 bucket, MQTT broker, and credential vault within 30 days. For immediate deletion email trust@tablize.com — we issue a deletion certificate. Backups roll over within 7 days, leaving no residual data.
DPAs, subprocessor list, pen-test results, SOC 2 status, Fly.io location attestation — email trust@tablize.com and we'll send a Trust Pack within 2 business days.
Request Trust Pack