Hirable: Your Personal AI Resume Tailor

Hirable is a two-phase agentic app that builds a rich profile of your professional background through a conversational interview, then generates a tailored resume and cover letter for any job you apply to — improving with every session.

How It Works

Phase 1 — Build your profile (do this once)

Upload your resume and Hirable will interview you to fill in what a CV can't capture: scale, impact metrics, private context per role, career goals. Your profile is saved to disk and enriched over time — the more detail you add, the better your tailored resumes become.

Phase 2 — Generate for each job

Paste a job URL or description. Choose:

• Quick mode — generate immediately from your profile
• Ask follow-ups mode — Hirable first identifies gaps between your profile and the JD (e.g. technologies mentioned in the job you haven't discussed), asks a few targeted questions, then generates

Output: a tailored resume (editable YAML + PDF) and a cover letter.

Features

• Persistent profile — built up incrementally through chat; never re-asks what you've already answered
• Private context — per-role notes and impact metrics inform the LLM but never appear on the PDF
• Parallel LLM adaptation — 6 resume sections adapted concurrently against the job description
• JD keyword highlighting — extracted keywords are bolded in the PDF
• Editable YAML — review and tweak the resume before rendering
• Professional PDF — rendered via RenderCV
• Multiple input formats — resume as PDF, DOCX, TXT, or TEX; job as URL or pasted text

Getting Started

1. Install dependencies

You'll need uv for package management.

# Install uv if you don't have it
curl -LsSf https://astral.sh/uv/install.sh | sh

uv sync
uv pip install "rendercv[full]"   # installed separately — strict deps conflict with uv resolver

2. Set your API key

Hirable works with OpenAI, Google Gemini, Anthropic, or any LangChain-supported provider.

Set the relevant environment variable:

export OPENAI_API_KEY="sk-..."       # OpenAI
export GOOGLE_API_KEY="..."          # Gemini
export ANTHROPIC_API_KEY="sk-ant-..." # Anthropic

Then edit src/config/__init__.py to select your provider and model names. The config has a small and large model slot — large is used for experience bullets and cover letter, small for everything else.

3. Run the app

streamlit run streamlit_app.py

4. Build your profile

Go to Profile Setup in the sidebar. Upload your resume (PDF, DOCX, TXT, or TEX) and click Load resume into profile. Hirable will automatically start a gap-analysis interview — answer as briefly or in as much detail as you like. Type done at any point to stop.

Your profile is saved to data/user_profile.json after every answer. You can re-open Profile Setup at any time to add more context.

5. Generate a tailored resume

Go to Generate Resume. Paste the job URL or description, choose Quick or Ask follow-ups mode, and click Start.

Once generated:

• Edit keywords to adjust bold highlighting in the PDF
• Expand the YAML editor to tweak any section
• Click Generate PDF to render and preview
• Download the YAML or PDF

Configuration

Setting	File	Description
Model provider & names	src/config/__init__.py	Switch between OpenAI / Gemini / Anthropic and set model IDs
Resume adaptation prompts	src/prompts/adaptation_prompts.py	Edit to tune resume output quality
Interview prompts	src/prompts/interview_prompts.py	Edit to change how gap analysis or Q&A extraction works
PDF style	src/config/rendercv_config.yaml	RenderCV theme and layout settings

Project Structure

hirable/
├── streamlit_app.py          # Home page
├── pages/
│   ├── 1_Profile_Setup.py    # Chat-based profile interview
│   └── 2_Generate_Resume.py  # Job input + generation UI
├── src/
│   ├── models/               # Pydantic data models
│   │   ├── user_profile.py   # Persistent profile (UserProfile, RoleContext, ProfilePatch)
│   │   ├── resume.py         # Output schema for PDF generation
│   │   ├── job_desc.py       # Parsed job description
│   │   └── cover_letter.py
│   ├── nodes/
│   │   ├── profile_nodes.py  # Gap analysis, interview Q&A extraction
│   │   └── generation_nodes.py # Resume adaptation, cover letter generation
│   ├── graphs/
│   │   └── generation_graph.py # LangGraph pipeline
│   ├── prompts/
│   │   ├── adaptation_prompts.py
│   │   └── interview_prompts.py
│   └── utils/
│       ├── persistence.py    # load/save user_profile.json
│       ├── export.py         # Resume → RenderCV → PDF
│       ├── parse.py          # File and URL parsing
│       └── llm.py            # LLM initialisation with rate limiting
└── data/
    └── user_profile.json     # Your persisted profile (gitignored)

Privacy

Your profile (data/user_profile.json) is stored locally only — nothing is sent anywhere except the LLM API calls you configure. Add data/ to your .gitignore if you're pushing to a public repo.