Built With
FastAPI
Next.js 16
Python
TypeScript
MongoDB🧠 LangGraph
⚡ Gemini 2.0
🔮 ChromaDB
Cold email automation. Google Form auto-fill. Multi-board job scraping. One agentic AI platform with a LangGraph pipeline and RAG system that turns your resume into personalised applications.
FastAPI Next.js 16 Python TypeScript MongoDBEvery repetitive step of your job search — automated, personalised, and intelligent.
Provide HR email + job description (text, screenshots, or PDFs). The LangGraph pipeline extracts details, retrieves your resume context via RAG, and drafts a personalised cover email — delivered via Resend API through your own verified domain (e.g., hi@yourdomain.com) with your resume PDF attached.
A Chrome Extension that scrapes Google Form questions, sends them to the Gemini backend, and auto-fills every field using your profile context — handling text, MCQ, checkboxes, and dropdowns intelligently.
Scrapes Indeed, LinkedIn, and ZipRecruiter in parallel. Two-layer deduplication (MD5 + MongoDB) ensures no duplicate listings. Results normalised into a unified schema with salary, location, and apply links.
Upload your resume, cover letters, and project documents. They're chunked, embedded with Gemini, and stored in ChromaDB. Every email and form answer is enriched with your most relevant experience — automatically.
Select a date range, fetch all past applications, and re-draft fresh cover emails with updated profile context. Concurrency-limited to 5 simultaneous LLM calls with exponential backoff retry.
Track every sent application, monitor delivery status, search by company/role, and view aggregate stats — all backed by MongoDB with real-time aggregation queries.
A 4-node LangGraph pipeline that transforms raw job postings into personalised, ready-to-send cover emails — delivered through your own custom domain via the Resend API, with your resume attached.
aryan@yourdomain.com) — not a generic sender. Resume PDF auto-attachedGemini Vision processes text + images → structured JSON with company, role, HR email, requirements
ChromaDB cosine similarity (0.93 threshold) + LLM cross-validation to flag recent duplicates
RAG retrieval: top-5 most relevant resume/project chunks from ChromaDB user_profile collection
Job info + profile context → Gemini drafts a compelling, personalised subject + body
Delivered through your own domain (e.g., aryan@yourdomain.com) — with resume PDF attached. Not a generic sender.
Injected into docs.google.com/forms/* — extracts questions, types, options, and form title
Service worker (MV3) sends structured question data to FastAPI backend
Extracts metadata, retrieves top-8 profile chunks, then single LLM call answers all questions
Extension fills each field. AnswerValidator ensures MCQ/checkbox answers match available options
A Manifest V3 Chrome Extension that scrapes Google Forms and auto-fills answers using your resume context — handling every question type intelligently.
Automated job discovery across three major job boards — scraped in parallel with robust deduplication and a unified listing dashboard.
User provides roles (e.g., "React Developer") + max results per role
python-jobspy queries Indeed, LinkedIn, ZipRecruiter simultaneously via asyncio.gather()
Internal MD5 hash dedup within run, then cross-run check against MongoDB scraped_jobs
New jobs inserted into MongoDB. All unique results returned to frontend with badges
A FastAPI modular monorepo with shared services, async MongoDB, and a React dashboard.
Clone the repo, add your Gemini + Resend API keys, upload your resume, and start sending personalised applications in minutes.