SimplicitLoading…
SimplicitSimplicit turns a frontend codebase (or a prompt) into a generated NestJS + Drizzle backend. This guide covers how it works end to end, what it produces, and where its limits are today.
Most backends for a given frontend are mechanical: the same auth boilerplate, the same CRUD routes, the same DTO validation, written by hand every time. Simplicit reads your frontend, works out what your backend needs to look like, and generates it — a real NestJS + Drizzle service on Postgres, not a prototype.
You stay in control at one specific point: the Entity Fields step, where you confirm exactly what data your backend will store before anything is generated. Everything after that is automated — generation, security review, stability checks, test writing, SDK generation, and deployment.
You give Simplicit your frontend; it analyzes the code, walks you through a short wizard to confirm what it found, runs the generation pipeline, and hands you a deployable backend with a working GitHub repository and live URL.
| Stage | What happens |
|---|---|
| Upload | Provide a ZIP of your frontend or connect a GitHub repository. |
| Analysis | Simplicit parses your codebase to detect entities, fields, relationships, and API calls. |
| Wizard | You review and correct what was detected — this is the Entity Fields step. |
| Pipeline | A sequence of generation stages builds, secures, tests, and packages the backend. |
| Deploy | Simplicit pushes the result to GitHub and deploys it to Railway. |
Analysis walks your codebase to build a picture of your app before anything is generated. It looks at:
This works best on React, Next.js, and Vue projects. The result feeds directly into the Entity Fields step — analysis proposes a data model, and you're the one who confirms it.
After analysis, you review the detected entities and their fields. You can add missing fields, correct types, and add entirely new entities.
This is the most important step — what you confirm here determines exactly what gets built: database columns, DTOs, validation rules, and routes. Nothing downstream is guessed twice; the pipeline builds strictly from what you approve here.
Once you confirm the Entity Fields step, generation runs as a sequence of stages, each responsible for one part of the backend:
| Stage | Responsibility |
|---|---|
| Architect | Turns your confirmed entities into a backend structure — modules, schema, and route plan. |
| Generator | Writes the NestJS modules, Drizzle schema, and DTOs. |
| Security | Reviews the generated code for common backend security issues before it ships. |
| Stability | Checks that the backend builds cleanly and holds together end to end. |
| Test Writer | Writes tests against the generated routes. |
| SDK Generator | Produces a typed client SDK for your frontend to call the new backend. |
Each stage has to complete before the next one starts, so a failure is always traceable to a specific step.
A NestJS + Drizzle ORM backend on PostgreSQL. Every generation includes an auth module, per-entity CRUD routes, typed DTOs with validation, a Drizzle schema, a typed SDK, and a health endpoint:
Your frontend talks to the new backend through the generated SDK rather than hand-written fetch calls:
Generated code isn't locked. Right-click any file in the IDE for:
You'll see a plain-English error naming the stage that failed and next steps. Click Retry to re-run all agents — you don't need to redo the Entity Fields step first.
This usually means the underlying AI provider hit a rate limit. It's non-fatal — the rest of the backend is complete and deployable. You can retry later to generate tests.
Add it manually in the Entity Fields step before continuing — anything you add there is treated the same as anything analysis detected on its own.
Double-check your Railway API token in Settings, then retry. If the token is valid and deploy still fails, the IDE error message will point to whichever step (repo creation, push, or Railway connection) didn't complete.
Is my code stored?
Your code is analyzed to design your backend, and your project files are saved to your account so you can view and edit them. It is never used for AI training.
Can I edit generated files?
Yes — right-click any generated file in the IDE for Explain, Edit, or Delete options.
What if the pipeline fails?
You'll see a plain-English error with next steps. Click Retry to re-run all agents.
Do I need a finished frontend before I start?
No. A rough prototype with the forms and pages you want is enough for analysis to work from — you can refine the frontend later.
What database does Simplicit use?
PostgreSQL, accessed through a Drizzle ORM schema that's generated alongside the rest of your backend.