Case studiesAI / productivity

FillWizard: AI-powered PDF form autofill

Full-stack product delivery for an AI PDF autofill tool: upload a form, detect fields (including OCR for scans), fill from a profile, and export a completed PDF.

Challenge

PDF forms are messy: AcroForm fields, flat scans, and inconsistent layouts. Users needed a path that felt reliable: detect fields, propose fills from a saved profile, let them review, then export, without drowning in engineering jargon.

Solution

We built the product around a clear review loop: upload, detect, map, approve, export. Frontend on Next.js App Router with Tailwind and shadcn/ui. Backend on NestJS with Prisma/PostgreSQL, BullMQ workers, S3-compatible storage, and pdf-lib / pdfjs for field handling. Polar handles billing.

Results

A shipping SaaS that turns form friction into a guided workflow, and a reusable pattern for document-heavy products: async jobs, human review before commit, and UI that stays calm when PDFs misbehave.

Outcomes

End-to-end PDF upload, detect, review, and export flow

Multilingual product surface (en / fr / ar with RTL)

Production NestJS + Prisma backend with async PDF jobs

Stack in practice

Monorepo with apps/web, apps/api, and shared packages. Next.js 16 on the client, NestJS + Prisma on the API, Redis/BullMQ for PDF work, and Polar as merchant of record. That same delivery shape shows up in our agency builds when clients need more than a brochure site.

Verified case studyLive product: www.fillwizard.com