Palak Narang

Chitkara University Punjab

2Projects

4Skills

1Achievements

Open to roles

Palak Narang

Featured project

Provenance: The Knowledge Graph That Understands Your Journey

Provenance transforms scattered certificates, resumes, and projects into a living, intelligent archive that doesn't just organize—it understands. Upload your credentials. The system reads them, classifies them automatically, embeds their meaning, and discovers the hidden connections between your skills, certifications, and experiences. A Python certification automatically links to your Python project. Your hackathon win surfaces when you search for "AI work." Your internship reveals the exact skills you brought to it. No manual tagging. No folder hunting. No guessing what you've accomplished. The Knowledge Graph visualizes your entire academic and professional journey as an interconnected web—every achievement, every skill, every opportunity—laid out in a way that makes your growth obvious. Search naturally ("show me my data science projects") and get answers instantly, not a list of filenames. This is what happens when RAG meets knowledge graphs meets a student's real problem: the system doesn't store documents. It understands people.

5 media files · github.com

View project

Core skills

PythonSQLpandasStatistics

MemoryVerse AI '26
PROVENANCE - REVIEWER NOTES QUICK START Backend setup: cd backend py -m venv venv venv\Scripts\activate pip install -r requirements.txt Create .env file, paste your OpenRouter API key python -m uvicorn app.main:app --reload --port 8000 Frontend setup (new terminal): cd frontend npm install npm run dev Open http://localhost:5173 Load demo data (optional, third terminal): cd backend venv\Scripts\activate cd .. python seed_demo_data.py This uploads 6 sample documents showing real relationships DESIGN PHILOSOPHY Forest-ink background with cream index cards (legible, premium feel) Brass/gold accents for key actions (hints at "filing" and "archiving") Typography: Fraunces serif for display, IBM Plex Mono for data No generic UI templates - custom CSS built for this product's identity WHAT TO LOOK FOR Upload a certificate mentioning "Python" Upload a project report also mentioning "Python" Click "Connections" on the project - AI automatically linked it to the cert with an explanation Scroll to "Knowledge Graph" - see the visual web of connections as nodes and edges Try searching: "show my AI projects" or "what skills have I used most" Check the timeline for chronological view TECHNICAL HIGHLIGHTS Module 1: Drag-drop upload with PDF/DOCX/image support (OCR via Tesseract) Module 2: LLM classification with forced JSON output (no parsing errors) Module 3: Local embeddings (sentence-transformers) + cosine similarity + LLM relationship labeling Module 4: Timeline rendering from dated documents Module 5: Real semantic search (embeddings, not keyword matching) Frontend: React with hand-crafted CSS, d3-force graph layout API ENDPOINTS GET /documents - list all uploaded docs GET /documents/timeline - docs sorted by date GET /graph - returns nodes+edges for knowledge graph visualization GET /search?q=... - semantic search GET /documents/{id}/related - connections for one doc POST /upload - upload a file WHAT'S NOT INCLUDED (future work) Certificate authenticity verification Multi-user support / authentication Cloud hosting (local/demo only) Natural language chat interface (could be added easily) DATABASE SQLite stored in backend/storage/digital_identity.db No external DB needed, runs anywhere API KEY Requires OpenRouter key (free tier available at openrouter.ai)
5 media files · github.comView

This is Palak’s work on Wooble.

Build a profile that shows what you can do — and share it anywhere.

Build yours

Palak Narang

Provenance: The Knowledge Graph That Understands Your Journey

Core skills

MemoryVerse AI '26