Tailr

AI-Powered Resume Intelligence Platform

Upload a resume, paste a job description — get an AI-optimized version grounded in your real experience.

Overview

Tailr is an AI-powered resume optimization platform. Unlike traditional resume generators, Tailr never fabricates experience — every optimization is grounded in the user's existing resume via RAG (Retrieval-Augmented Generation).

Core flow:

1. Upload a resume (PDF/DOCX/TXT) → LlamaExtract extracts structured JSON
2. Paste a job description → LlamaExtract extracts structured requirements
3. Section-by-section embeddings → semantic search in Qdrant Cloud
4. Retrieve context → LLM optimizes each section
5. Display the optimized result

Architecture

flowchart TD
    A["Resume File<br/>(PDF / DOCX / TXT)"] --> B["LlamaExtract<br/>(Structured JSON)"]

    B --> C["Section Embeddings<br/>(Qdrant Cloud)"]
    B --> D["Job Description<br/>(LlamaExtract)"]

    C --> E["Semantic Retrieval"]
    D --> E

    E --> F["Gemini 2.5 Flash LLM<br/>(LangGraph Workflow)"]
    F --> G["Optimized Resume Output"]

Loading

Features

• AI-Powered Extraction — LlamaExtract converts PDF/DOCX/TXT into structured JSON (experience, education, skills, projects, certifications, achievements)
• Section-by-Section Embeddings — Each section (summary, skills, experience, etc.) is independently embedded for precise retrieval
• Semantic RAG — Qdrant Cloud vector search retrieves the most relevant sections for the target job
• Multi-Agent Workflow — LangGraph orchestrates JD analysis, resume analysis, planning, rewriting, and optimization
• Streaming Results — Frontend receives real-time SSE events as the workflow progresses
• Data Management — View, use, or delete previously extracted resumes and job descriptions
• JWT Authentication — Secure user accounts with bcrypt password hashing
• Clean Architecture — Domain-driven design with repository pattern and dependency injection

Tech Stack

Frontend

• Next.js 16 (App Router)
• React 19
• TypeScript (strict mode)
• Tailwind CSS v4
• React Query v5
• Zustand
• ShadCN UI

Backend

• Python 3.13
• FastAPI
• SQLAlchemy 2.x (async)
• Alembic
• Pydantic v2
• LangChain
• LangGraph

AI & Vector

• Google Gemini 2.5 Flash (LLM)
• Google Gemini Embeddings (models/gemini-embedding-001, 3072 dims)
• LlamaExtract (structured data extraction from files)
• Qdrant Cloud (vector database)

Storage

• PostgreSQL 17 (relational data)
• Qdrant Cloud (vector embeddings)
• Redis (caching — optional)

Getting Started

Prerequisites

• Python 3.13+
• Node.js 22+
• PostgreSQL 17
• Qdrant Cloud account
• Google Gemini API key
• LlamaExtract API key

Backend

cd apps/backend
python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt

Create .env in apps/backend:

DATABASE_URL=postgresql+asyncpg://postgres:admin@localhost:5432/tailr
GEMINI_API_KEY=your_key_here
LLAMAEXTRACT_API_TOKEN=your_token_here
QDRANT_URL=https://your-instance.eu-west-1-0.aws.cloud.qdrant.io
QDRANT_API_KEY=your_key_here
JWT_SECRET_KEY=your_secret

Run migrations:

alembic upgrade head

Start the server:

uvicorn api.main:app --reload

Frontend

cd apps/frontend
npm install
npm run dev

Project Structure

apps/
├── backend/
│   ├── api/
│   │   ├── routes/          # FastAPI route handlers
│   │   └── main.py          # App entrypoint
│   ├── application/         # Use cases / service layer
│   ├── domain/              # Business entities & repository interfaces
│   ├── infrastructure/
│   │   ├── database/        # SQLAlchemy models & session
│   │   ├── llamaindex/      # LlamaExtract + Qdrant vector store
│   │   └── repositories/    # SQLAlchemy repository implementations
│   ├── workflows/           # LangGraph workflow nodes & graph
│   └── alembic/             # Database migrations
└── frontend/
    ├── app/                 # Next.js App Router pages
    ├── components/          # React components
    └── lib/                 # Zustand stores & utilities

Development Status

Module	Status
User Auth (JWT)	✅ Complete
Resume Upload & Extract	✅ Complete
JD Upload & Extract	✅ Complete
Qdrant Indexing	✅ Complete
Semantic Retrieval	✅ Complete
LangGraph Workflow	✅ Complete
SSE Streaming	✅ Complete
Data Management	✅ Complete
Guardrails	⏳ Planned
ATS Scoring	⏳ Planned
PDF Generation	⏳ Planned

License

MIT

Built with FastAPI, Next.js, LangGraph, Gemini, Qdrant Cloud, and LlamaExtract.