SANJANA MAHMOOD
Hi, I'm Sanjana.
This is how I usually
appear in the world!
About me
I'm a Full-Stack Developer and Generative AI Engineer, passionate about building intelligent, scalable systems.
I specialize in architecting end-to-end applications built on JavaScript and Python ecosystems, leveraging frameworks like React, Node.js, and FastAPI alongside PostgreSQL and MongoDB databases. With hands-on experience across AI-powered platforms, data privacy solutions, and blockchain integrations, I design systems that are secure, performant, and cloud-ready, using Docker, GitHub Actions, and agile workflows.
Today, I focus on AI agent development, LLM integration, and end-to-end full-stack systems that empower teams to automate complex workflows, enhance user intelligence, and scale efficiently.
Technical Skills
Languages
Web & APIs
Database
AI & Generative Systems
Tools and Cloud
Experience
Software Developer
GaiaMine Limited
April 2025-Present
- Collaborated in developing a tokenized investment platform using React, TypeScript, Node.js, and MongoDB, supporting 99%+ production uptime.
- Created Docker-based CI/CD pipelines using GitHub Actions, reducing deployment errors by ~60%.
- Built a RAG pipeline (LangChain, FAISS, Gemini 2.5 Flash) to auto-generate Mining Company Initiation Reports, cutting analyst prep time by 90%+.
Cyber Risk Analyst
Deloitte Advisory (Hong Kong) Limited
Sept 2023 – Sept 2024
- Developed a Data Privacy assessment platform using React, Survey.js, Node.js, PostgreSQL to automate client compliance reviews.
- Optimized backend queries and refactored codebase, reducing SQL execution time and improving performance by 15%.
- Implemented and monitored Forcepoint DLP solution to detect and prevent data leakage across managed client environments.
- Optimized infrastructure monitoring with Azure Monitor, achieving 30% faster response time and 99.9% uptime.
- Automated DLP processes using Python and PowerShell, reducing manual efforts by 50%.
Research Assistant (Internship)
The Chinese University of Hong Kong
June 2022-Aug 2022
- Fine-tuned a real-time human pose estimation model (MMPose) and built a PyTorch inference pipeline for human aerosol spread prediction from RGB-D data.
- Built a 3D reconstruction service with Open3D, reducing visualization latency by 35% via asynchronous streaming.
Key Projects
PDF RAG Application
A full-stack application that enables AI-powered conversations with PDF documents. This project leverages Azure OpenAI's GPT-4 and embedding models to provide intelligent, context-aware responses based on uploaded PDF content.
AI Resume Analyser
An intelligent resume analysis tool that uses natural language processing to extract key information, evaluate skills, and provide personalized feedback on resumes. It helps job seekers improve their resumes and assists recruiters in candidate screening.
URL Shortener
A full-stack URL shortening platform with comprehensive analytics and QR code generation. Features real-time click tracking, geographic location insights, device statistics, and an intuitive dashboard for managing shortened links with custom aliases.
Eveagle Academy Platform
Built a comprehensive web application using the MERN stack for an EdTech startup in Hong Kong. Features course management, student enrollment, and an intuitive user interface designed for streamlining educational operations.
Reversi Game
A classic two-player strategy board game implemented in Java with an intuitive GUI. Features turn-based gameplay, valid move validation, and win detection. Players compete to control more pieces by flanking and capturing opponents' discs.
Education
B.Sc. in Computer Science
The Chinese University of Hong Kong
September 2019 - July 2023