Case study · Esketchers · 2024–2025
Esketchers Proposal Generator
An internal Django RAG that matches job descriptions to company portfolios — 90%+ relevance via hybrid semantic + keyword search.
- Django
- RAG
- pgvector
- AWS
- Nginx/Gunicorn
- CI/CD
90%+job-description-to-portfolio match relevance
Internal company tool — case study only; no public link or repository.
The problem
Writing proposals meant manually hunting for relevant past work across a large company portfolio.
My role
Senior Software Engineer — designed and deployed the RAG system.
What I built
A Django RAG document-intelligence platform with pgvector embeddings, AI-generated summaries, and multi-format (PDF/DOCX) processing, using hybrid semantic + keyword search to match job descriptions to portfolio projects. Deployed on AWS (EC2, RDS PostgreSQL + pgvector) with Nginx/Gunicorn and CI/CD.
Architecture & stack
- Django RAG over company portfolios
- pgvector embeddings + hybrid semantic/keyword search
- Multi-format document processing (PDF/DOCX); AI-generated summaries
- AWS EC2 + RDS (PostgreSQL/pgvector), Nginx, Gunicorn, CI/CD