Frequently asked questions
The questions engineering and security teams ask most, answered plainly. Don't see yours? Ask us — we answer honestly.
Getting started: first login
- How do I log into Workbench the first time — where is my first-run setup token?
- After your Professional Cloud (AWS Marketplace) deployment is running, open Workbench over HTTPS. On a fresh install it shows a first-run bootstrap screen that asks for a one-time setup token — there is no default password. Read the token from the running deployment with the command below (the guided installer also prints it when it finishes), paste it into the bootstrap screen, and create your administrator account. From then on you sign in normally. The token is single-use and held in memory, so it changes if the pod restarts before the admin account exists; re-run the command to read the current one. Treat it as a credential.First login & your setup token →
kubectl -n elevarq logs deploy/elevarq-elevarq-unified | grep "first-run setup token"
The AI and your data
- Does Elevarq use AI — and will it hallucinate wrong advice?
- The findings themselves come from deterministic detection rules, not AI. Elevarq Insight uses local models to explain each finding in plain language, and every explanation is validated against the finding's cited evidence — anything that goes beyond it is discarded and the deterministic explanation is shown instead. The model explains; it never decides, invents, or inflates. How we keep the AI honest →
- Which model does it use? Is it really local?
- Local models trained specifically for PostgreSQL (and TimescaleDB on Professional and above), running inside your environment — not a general-purpose cloud LLM pointed at your database.
- Does my data leave my network or train your models?
- No. Elevarq is offline-activated and runs entirely in your infrastructure. Snapshots, query content, findings, and tickets stay in your network and never train Elevarq's models.
- Are there token costs or usage caps?
- No. Because the model is local, there is no per-token billing and no usage cap — a busy database costs the same as a quiet one.
Running Elevarq
- Do I need a GPU?
- Starter runs on CPU. Professional and above run on a GPU, which powers the larger models. You can start on CPU and move up as your needs grow. Capabilities by plan →
- Where does it run, and what does it touch?
- Everything runs in your infrastructure as containers. Elevarq Signals is the only component that connects to PostgreSQL, and only with a read-only role on a schedule you control. Analyzer and Insight work off-production: the analysis runs in an isolated Report Database, so no customer query runs against your live system and no query-execution load is added to it. How it works →
- What database access does it need?
- A read-only monitoring role (pg_monitor-level privileges). No writes, no schema changes, and no agent on the database host.
- How long until I see value?
- Setup takes minutes. The first run surfaces the obvious wins; because Elevarq analyzes every run, it builds context over time and catches slow regressions and configuration drift that a single look would miss.
Compatibility
- Does it work with Amazon RDS or other managed PostgreSQL?
- Yes. Elevarq works with standard PostgreSQL wherever you run it — self-managed or managed (Amazon RDS, Google Cloud SQL, Azure Database for PostgreSQL). The only requirement is a database that behaves like standard PostgreSQL. Compatibility →
- Do you support TimescaleDB and pgvector?
- TimescaleDB is supported on Professional and above, with extension-aware analysis. pgvector is on the roadmap and not included yet.
What it does
- What problems does it actually detect?
- Elevarq detects developing conditions across vacuum and bloat, indexing, query performance, configuration and memory, WAL and replication, connections and locking, schema and data model, security, and TimescaleDB. What Elevarq detects →
- Does it change my database?
- No. Elevarq recommends and writes an implementation-ready ticket to your tracker; a human reviews and ships every change through your normal review and CI/CD process. Nothing is ever run against your database automatically.
- Which ticket systems does it integrate with?
- GitHub, GitLab, Jira, and Linear.
Security and compliance
- Are you SOC 2 or ISO 27001 certified?
- Elevarq's engineering is aligned with SOC 2 and ISO 27001 practices and is designed for auditability; we do not claim certification. Because Elevarq runs entirely in your environment and never receives your data, most of the data-handling surface stays under your control. Security →
- Can it run air-gapped?
- Yes, with the boundaries made explicit. Once the container images and the offline activation artifact are provisioned, core Analyzer operation needs no outbound connectivity — analysis runs entirely inside your perimeter, with no telemetry, no phone-home, and no cloud inference. Two things use the network, both under your control: installation (and updates) may pull images from a container registry, and any ticket integrations you enable connect only to the tracker you choose and are optional. For fully offline operation, Enterprise includes packaged air-gapped update artifacts, so you can update without the registry pull. Security →
Pricing
- How is Elevarq priced?
- Four plans — Starter, Professional, Business, and Enterprise — with annual (about 2 months free) or monthly billing. You get started by contacting our team, and Enterprise is custom and sales-led. There is no token billing and no usage cap. Pricing →
Still have a question?
Tell us what you run and what you're trying to solve — we answer honestly, whether or not Elevarq is the right fit.