AWS Credits for Startups: How AWS Activate Works and How to Make Credits Last

AWS credits are one of the few startup perks that can cover most of a company's infrastructure bill for its first year or two, and they are easy to waste. Credits spent on an oversized database, a NAT gateway nobody needed, or a staging environment that runs around the clock buy nothing, and when they run out the bill continues at the same rate. This guide covers how AWS Activate works today, how to apply, and how to set up a backend so credits pay for the product rather than for idle resources.

AWS Activate credit tiers

AWS runs its startup credits through AWS Activate, with tiers based on funding stage, as listed on the AWS Activate credits page:

Activate Providers include accelerators, angel groups and venture firms, and if an investor or accelerator is a provider, it gives you an Organization ID to apply with. Credits are in US dollars and subject to the AWS Promotional Credit Terms and Conditions, which set the expiry date and the services they apply to, so read them before planning around a balance.

Eligibility

AWS lists these requirements for Activate:

• Pre-Series B.
• Founded in the last 10 years.
• New to Activate credits, or requesting more credits than previously received.
• An AWS account linked to the application.

How to apply

1. Create an AWS Builder ID with a professional email address and complete the profile.
2. Pick the tier: self-funded companies apply to Founders, and funded companies check whether their investor or accelerator is an Activate Provider and get its Org ID for Portfolio.
3. Describe the startup, including product, funding stage and company details.
4. Link or create the AWS account and submit the application, and AWS says to expect an answer within 5 to 10 business days.

A company website and a professional email domain help, since the application asks for company details AWS uses to confirm eligibility.

Where startup credits usually go

For a typical SaaS backend, most of the bill comes from a few places:

• Compute: ECS on Fargate, EC2 or Lambda.
• Databases: RDS or Aurora, usually the largest line once Multi-AZ is on.
• Networking: NAT gateways, which charge per hour and per GB processed, load balancers and data transfer.
• AI workloads: Amazon Bedrock model usage and GPU instances for AI startups.

How to make AWS credits last

Size for today's traffic. A small database instance and two small Fargate tasks carry most early products. Scale up when metrics show the need.

Turn off what does not serve users. Development and staging environments do not need production-sized instances or Multi-AZ. Tear down preview stacks after merge.

Watch the networking line. NAT gateways and cross-AZ traffic are a common surprise on early AWS bills. Check Cost Explorer by service in the first weeks.

Set budgets and alerts. AWS Budgets can alert when forecasted spend passes a threshold, which tells you months ahead when credits will run out.

Avoid paying for operations with engineering time. Credits cover AWS usage, not the weeks spent writing Terraform, IAM policies and CI pipelines. For a small team, that time is usually the larger cost.

Spending credits on a backend in your own AWS account with Encore

Encore is infrastructure from code for TypeScript and Go, with a platform that provisions and deploys into your own AWS account. The database, queues, cron jobs and object storage a backend uses are declared in its code, and Encore creates them as RDS, SNS and SQS, and S3 in the account where your Activate credits apply, without Terraform. Encore provisions an IAM role for the service, and you keep the account and its resources if you ever stop using Encore.

A signup service with a database and a queue for welcome emails:

// signup/signup.ts
import { api } from "encore.dev/api";
import { SQLDatabase } from "encore.dev/storage/sqldb";
import { Topic, Subscription } from "encore.dev/pubsub";

const db = new SQLDatabase("signup", { migrations: "./migrations" });

interface SignedUp {
  userId: number;
  email: string;
}

export const signups = new Topic<SignedUp>("signups", {
  deliveryGuarantee: "at-least-once",
});

export const register = api(
  { expose: true, method: "POST", path: "/signup" },
  async ({ email }: { email: string }): Promise<{ userId: number }> => {
    const row = await db.queryRow<{ id: number }>`
      INSERT INTO users (email) VALUES (${email}) RETURNING id
    `;
    await signups.publish({ userId: row!.id, email });
    return { userId: row!.id };
  },
);

const _ = new Subscription(signups, "send-welcome-email", {
  handler: async (event) => {
    // Send the welcome email for event.email here.
  },
});

The migration in signup/migrations/1_create_users.up.sql creates the table:

CREATE TABLE users (
    id SERIAL PRIMARY KEY,
    email TEXT NOT NULL UNIQUE
);

encore run starts Postgres and the Pub/Sub topic locally, so development does not spend credits at all. In the dashboard you connect your AWS account and create environments, and each environment can run all services in one process or each service separately, which keeps a development environment small while production scales, as covered in the environments docs. On AWS, production databases are RDS with Multi-AZ and backups, described in the AWS infrastructure docs. Every app also starts with a staging development environment on Encore's free development hosting, so that environment does not draw on your AWS credits.

A sensible order

1. Apply for Activate Founders now if you are self-funded, or ask your investor or accelerator for its Org ID.
2. Set up the AWS account with budgets and alerts on day one.
3. Run development locally and staging off your AWS account, or at minimum on small instances.
4. Size production for current traffic and review Cost Explorer monthly.

Related guides

• Deploy to AWS without Terraform
• Deploy to AWS without Kubernetes
• What is Amazon ECS
• Best tools for AI agents to deploy to AWS