Avani Goswami

I’m an AI/ML-focused Computer Science Engineering student building production-minded systems across NLP, LLM integration, RAG, and agentic AI. I enjoy translating complex requirements into scalable pipelines that actually run in real workflows.

At PrragetX S Softwares Private Limited, I developed a resume evaluation workflow that parses resumes, matches them to job descriptions, and generates alignment scores across multiple metrics. I also built an OCR-based document processing pipeline that extracts structured data from images and PDFs, then stores outputs to Supabase and Google Sheets.

I’ve built backend services in Python with FastAPI—including deal and coupon management logic—and I enjoy wiring AI systems into practical APIs with authentication and database layers. I’ve also created messaging automation bots for WhatsApp and Instagram to support workflow-based actions.

My projects reflect how I combine computer vision, LLMs, and retrieval: I built a real-time surveillance event intelligence system using YOLOv8, DeepSORT, OpenCV, and a FAISS-powered RAG layer with Groq API. I also designed an agentic productivity co-pilot pipeline deployed with a Streamlit dashboard, and I’ve worked on CNN + RAG and NLP sentiment pipelines to deliver measurable accuracy and deployment-ready demos.