What is the difference between the tiers?

The two tiers are genuinely additive. Flat Multi-User is the base: standard user roles, platform invitations, a global AI gateway, and admin controls. Multi-Tenant & Stripe includes everything in Flat Multi-User, plus organization workspaces, seat limits, Stripe subscription billing (Free/Pro/Enterprise), customer portals, credit top-ups, and per-organization API keys with global fallback.

Does this product include an automated test suite?

No — both tiers ship a configured pytest harness (pytest.ini, and pytest/pytest-asyncio/httpx as dev dependencies) but no actual test files. You get the setup, not a written suite. We're telling you this upfront rather than have you find out after purchase.

Can I run the system entirely in Docker?

Yes. BoilerShelf includes a pre-configured docker-compose.yml file. Running 'docker compose up --build' starts the FastAPI server, a Redis broker, a PostgreSQL instance, and the Celery background worker concurrently, making deployment simple.

Which LLM model providers does the gateway support?

The gateway is integrated with OpenAI (GPT models), Anthropic (Claude models), Google Gemini, and Groq (Llama models). It handles upstream rate limits automatically and calculates costs based on active model pricing tables.

How are provider API keys secured?

Keys are encrypted prior to database storage using AES-256 (via the cryptography Fernet package) utilizing a unique server-side ENCRYPTION_KEY. Plain API credentials are never written to database tables or exposed in API response structures.

Multi-Tenant tier — Dashboard: tenant workspace overview with token quota, spend projection, and team seats

1 / 12

Micro-SaaS Productsadvanced

BoilerShelf AI SaaS Core

Full B2B SaaS architecture with isolated workspaces and Stripe payments

$179one-time payment

Everything in Flat Multi-User Platform, plus:
Multi-tenant workspace isolation (organizations)
Automated Stripe subscriptions (Free, Pro, Enterprise plans)
Stripe Customer Portal redirect and webhook syncing
One-Time Credit Top-Ups with admin-configurable conversion rates
Workspace-level API key isolation (BYOK) with global fallback
Quota and seat limit validations on active organization plans
Organization suspend/unsuspend and member management (24 admin routes vs. 8 in Flat Multi-User)
Next.js organization switcher and billing/plans UI

Instant download

Secure checkout

Refund if we can't fix it

Commercial license

Response in 24h

Have a question before buying? Contact us

10 minSetup Time

80 hrsDev Time Saved

20Pages

At a glance

Multi-tenant AI applications with organization-level billingB2B SaaS with team seat limitationsUnified AI proxy routing with cost calculationsInternal tools or single-team AI products that don't need multi-tenancy

Setup in about 10 minutes

Saves roughly 80 dev hours

Infra estimate: $0–$10/month on free/low-cost tiers of Postgres and Redis, plus pay-as-you-go LLM API usage

Built with

Next.jsReactTypeScriptTailwind CSSZustandTanStack QueryAxiosFastAPI

Product details

About this product

BoilerShelf AI SaaS Core

Start an AI SaaS on a proven FastAPI + Next.js foundation without rebuilding the production plumbing. BoilerShelf AI SaaS Core is a production-grade, full-stack starting point built with FastAPI, PostgreSQL, Redis, and Celery on the backend, and Next.js + TypeScript on the frontend, providing multi-user authentication, secure credential storage, and a unified AI provider gateway with cost tracking.

Core capabilities

Unified AI Gateway & Routing: Proxy requests to OpenAI, Anthropic, Gemini, and Groq with automatic retry logic, rate limit handling, and cost calculation.
Role-Based Access Control (RBAC): Flexible user roles and permissions enforcing fine-grained endpoint authorization.
Asynchronous Task Queue: Celery + Redis worker integration for non-blocking usage logging, calculations, and diagnostics.
Enterprise Operations & Diagnostics: Live system statistics, Sentry error tracking, structured JSON logging, and correlation ID tracing.

Two tiers, one additive foundation

Unlike some of our other multi-tier products, these two tiers are genuinely additive — every API route in the Flat Multi-User tier also exists in the Multi-Tenant tier, verified directly against both codebases.

Multi-Tenant & Stripe Billing: The full B2B SaaS architecture. Isolated organization workspaces, seat limits, Stripe subscription plans (Free, Pro, Enterprise), one-time credit top-ups, and organization-level API keys with global fallbacks — everything in Flat Multi-User, plus organizations and billing.
Flat Multi-User Platform: A streamlined, user-centric system. Standard roles (superadmin, admin, user), secure email invitations, global AI key management, and administrator controls (user directory, account banning) — the same AI gateway and diagnostics, without multi-tenancy or billing overhead.

An honest note on testing

Both tiers ship a configured pytest harness (pytest.ini, test dependencies in pyproject.toml) but no actual test files — you're getting the setup, not a written test suite. We'd rather tell you this upfront than have you discover it after purchase.

fastapinextjsboilerplatesaasmulti-tenancystripeceleryai-gatewaypythonmulti-tier

Key features

Unified AI Gateway Routing

Proxy prompts to OpenAI, Anthropic, Gemini, or Groq using a single clean endpoint, complete with automated upstream rate-limit retries.

Symmetric Credential Encryption

Encrypt API keys at rest using AES-256 Fernet symmetric encryption, ensuring credentials are never exposed in logs or plain responses.

Celery & Redis Task Worker

Process cost calculations, prompt auditing, and usage tracking asynchronously to preserve low latency for active AI sessions.

Double-Authenticated User Auth

Provides secure email OTP verification, bcrypt password hashing, login browser/device monitoring, and secure HttpOnly cookie refresh token rotation.

Streamed CSV Usage Exports

Generate and stream real-time CSV downloads of monthly AI token usage and costs directly to administrators.

Interactive OpenAPI Specs

Auto-generated and fully interactive Swagger/ReDoc endpoints available instantly on server startup for fast API testing.

What's included & Fit

Included

FastAPI backend application source code with fully implemented routers and schemas

Next.js + TypeScript frontend dashboard application source code, wired to the backend API

Celery background worker logic and Redis cache configurations

Alembic database migrations setup with preconfigured model relationships

Database seeding script preloaded with users, organizations, model pricing, and logs

Multi-stage Dockerfile and docker-compose.yml configuration for seamless deployment

Clean Markdown documentation covering architecture, database design, and setup guides

Not included

Automated test suite — pytest is configured (pytest.ini, test dependencies in pyproject.toml) but no test files ship; you write your own

Pre-configured live Stripe dashboard accounts or credentials

Sentry error logging endpoints or email dispatch SMTP servers

Target audience

Technical founders, Python developers, or small agencies starting a greenfield AI product

SaaS founders launching B2B or B2C AI applications

Developers needing secure credential management and AI proxy routing

Perfect for

Multi-tenant AI applications with organization-level billing

B2B SaaS with team seat limitations

Unified AI proxy routing with cost calculations

Internal tools or single-team AI products that don't need multi-tenancy

Not suitable for

Projects requiring NoSQL databases out of the box

Non-Python backend or non-Next.js frontend tech stacks

Teams that need a ready-made automated test suite (see the testing note in the description)

BoilerShelf AI SaaS Core · Multi-Tenant & Stripe Billing

$179

one-time payment

Instant download Secure checkout Refund if we can't fix it Commercial license Response in 24h

Details

commercial License

Email Support

Languages

Next.jsReactTypeScriptTailwind CSSZustandTanStack QueryAxiosFastAPIPythonCeleryRedisPostgreSQLSQLAlchemyAlembic

Status

Version1.0.0

Last updatedNov 2026

Impact

80+ hours saved

Estimated dev time

~10 min setup

Time to first run

$179