> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tensorcost.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Get started

> Sign up, connect your first cloud account, see your first recommendation within 48 hours.

# Get started

This guide takes you from zero to your first verified recommendation. Three steps, and the activation event — your first accepted recommendation — is typically reached within 48 hours of connecting your first account.

<Note>
  TensorCost is built around three workload classes — **GPU fleets**, **managed inference** (Bedrock and friends), and **agent workloads**. You don't have to connect all three to get value; most customers start with whichever one is dominating their bill.
</Note>

## Prerequisites

* A TensorCost workspace. Sign up at [tensorcost.com](https://tensorcost.com) or accept your design-partner invite.
* One of:
  * An AWS account with Bedrock usage (or CUR 2.0 enabled), **or**
  * An Azure subscription with Azure OpenAI usage, **or**
  * A GCP project with Vertex AI usage, **or**
  * GPU instances on AWS / Azure / GCP / Kubernetes / bare-metal that you can run a Docker container on.
* An admin (to mint credentials and connect cloud accounts) and, for the GPU agent, someone who can run a Docker container on the target host.

## Step 1 — Create your tenant and invite your team

<Steps>
  <Step title="Sign up">
    Visit [tensorcost.com](https://tensorcost.com) or follow the design-partner invite email. Cognito-backed sign-up; SSO is enabled via your tenant admin once the tenant is provisioned.
  </Step>

  <Step title="Invite teammates">
    From the shell sidebar, open **Settings → Members**. Three roles ship by default:

    | Role     | What they can do                                                                                 |
    | -------- | ------------------------------------------------------------------------------------------------ |
    | `member` | Read dashboards, accept/dismiss recommendations, see their own team's spend.                     |
    | `admin`  | All of `member` plus connect cloud accounts, manage agents, configure alert routes, set budgets. |
    | `owner`  | All of `admin` plus billing, tenant deletion, RBAC changes.                                      |
  </Step>

  <Step title="Map your tags (optional but recommended)">
    Open **Settings → Tag mapping** and bind your existing AWS/Azure/GCP cost-allocation tags to the TensorCost dimensions: `application`, `team`, `environment`, `owner`. This is what powers attribution; without it, everything rolls up under "untagged."
  </Step>
</Steps>

## Step 2 — Connect your first source

Pick the path that matches what's burning the most money first. You can layer in the others later.

### Path A — Amazon Bedrock (lead managed-inference path)

This is the fastest path to a first recommendation because it requires no agent install.

<Steps>
  <Step title="Open the Bedrock wizard">
    **Integrations → Add AWS Bedrock**. The wizard auto-suggests an `ExternalId` — accept it.
  </Step>

  <Step title="Choose onboarding mode">
    `SingleAccount` (default — one AWS account) or `Organization` (consolidated billing with payer + member-account jump roles). Most early customers run `SingleAccount`. See [bedrock integration](/bedrock-integration) for the multi-account variant.
  </Step>

  <Step title="Enable Bedrock model-invocation logging">
    AWS console → Bedrock → Settings → Model invocation logging → CloudWatch destination. Note the log-group ARN.

    <Warning>
      The log group must be in the same region as your `InvokeModel` calls. Logging in `us-west-2` while your traffic runs in `us-east-1` is the most common day-1 silent failure.
    </Warning>
  </Step>

  <Step title="Deploy the CloudFormation stack">
    Click the one-click CFN link in the wizard. The stack creates exactly one IAM role (`TensorCost-BedrockReader-<ExternalId>`) with read-only Bedrock + CloudWatch permissions and an external-ID-bound trust policy. Nothing else.
  </Step>

  <Step title="Validate the connection">
    Paste the role ARN back into the wizard and click **Validate**. The wizard polls STS-AssumeRole + a sample CloudWatch read for up to two minutes.
  </Step>
</Steps>

Your dashboard backfills 90 days of CUR + CloudWatch data within 30–60 minutes. Within 48 hours, the four MVP recommenders surface routing, prompt-cache, provisioned-throughput, and runaway-loop findings with \$-impact estimates.

### Path B — Install the unified GPU agent

For GPU fleets running on EC2, EKS/GKE/AKS, on-prem Slurm, or Ray. Full guide in [agent installation](/agent-installation).

**Integrations → Agents → Add agent** mints per-agent credentials and gives you a ready-to-run command:

```bash theme={null}
docker run -d \
  --name tensorcost-agent \
  --restart=unless-stopped \
  --gpus all \
  -e BACKEND_API_URL='https://api.tensorcost.com' \
  -e TENANT_ID='<from the console>' \
  -e AGENT_HOSTNAME='<from the console>' \
  -e AGENT_API_KEY='<from the console>' \
  public.ecr.aws/g2c6m0v1/agent:latest
```

The agent auto-detects EC2 / GKE / AKS metadata via IMDSv2 where available and starts reporting over plain HTTPS. Metrics start flowing within a few minutes of startup. A long-lived gRPC transport (lower latency, needed for remote commands) is also available as an opt-in — see [agent installation](/agent-installation#transport-modes) for the trade-offs.

### Path C — Azure OpenAI / Vertex / OpenAI API / Anthropic API

Same pattern as Bedrock, with provider-specific credentials. **Integrations → Add provider →** pick the source. Each adapter ingests:

* Per-request: model, input tokens, output tokens, latency, cache-hit rate
* Daily billing: cost normalized to `ai_spend_events`
* Tags / metadata: mapped to your `application` / `team` / `environment` / `owner`

Raw prompts and responses are **never** stored. Hashes only. See [SOC 2 readiness](/soc2-readiness-guide#redaction-at-ingestion) for the redaction guarantee.

### Path D — AWS / Azure / GCP infrastructure cost

For general cloud spend beyond AI usage — compute, storage, networking, everything else on your bill. This is a separate connection from Paths A–C above, even against the same cloud account. **Build → Connect**, Cloud tab → pick a provider. AWS is a cross-account IAM role (a few minutes); Azure is secretless but needs a directory admin and a subscription owner (\~10 minutes); GCP is a service account key or Workload Identity Federation pointed at your BigQuery billing export (10–15 minutes). Full walkthrough for each: [connect cloud accounts](/connect-cloud-accounts).

## Step 3 — See your first recommendation

Within 48 hours of connecting your first source, the **Recommendations** feed populates. Each entry includes:

* A specific, dollar-quantified change ("route 14% of `customer-support-agent` traffic from Claude Opus 4.6 to Haiku 4.5 — \$4,200/month").
* The evidence (sample request IDs, cost breakdown, A/B plan).
* Accept / dismiss-with-reason / snooze actions.

Acceptance is the activation event. Once you accept, the savings ledger starts tracking realized savings against the baseline. Verified savings populate after a 30-day window.

## What to do next

<Columns cols={2}>
  <Card title="Set up alert routes" icon="bell" href="/configuration#notification-channels">
    Slack, PagerDuty, email, Microsoft Teams, custom webhook.
  </Card>

  <Card title="Define budget hierarchies" icon="coins" href="/configuration#budgets-and-rbac">
    Tenant → team → application. Burn-rate alerts at 50%, 80%, 100%.
  </Card>

  <Card title="Connect your second source" icon="layer-group" href="/bedrock-integration">
    Coverage compounds. Customers with all three workload classes see 2× the recommendations.
  </Card>

  <Card title="Wire up MCP" icon="terminal" href="/realtime-events#mcp">
    Query TensorCost from Claude Desktop or your own agents.
  </Card>
</Columns>

## When you get stuck

* Check the **Sync history** drawer on the connection — every error from STS, CloudWatch, or the IAM trust policy surfaces here with a remediation link.
* Common day-1 failures (and remediations) are catalogued in our [customer onboarding runbook](/bedrock-integration#top-day-1-failures).
* Email [support@tensorcost.com](mailto:support@tensorcost.com) — design partners get a shared Slack Connect channel.