Saakshi Gupta

Saakshi Gupta

AI Enthusiast

Adelaide, Australia

About Me

$ whoami|


I'm a AI Enthusiast focused on LLM agents, RAG and multimodal AI. I build scalable services and multi-agent architectures that automate workflows and deliver measurable impact.

I hold a Bachelor's Degree in Software Engineering with a Data Science minor from the University of Queensland, where I was a recipient of the UQ Global Connect Scholarship.

My work spans deepfake detection research, generative AI systems, and multimodal content analysis. I'm passionate about building AI systems that are both powerful and explainable.

Python Python
Java Java
JavaScript JavaScript
C/C++ C/C++
React React
Node.js Node.js
TypeScript TypeScript
CSS CSS
AWS AWS
S3, Nova API, Lambda
Git Git
Linux Linux
Bash Bash

ML & AI

🔥
PyTorch
🧠
TensorFlow
🤗
Hugging Face
🖼️
CNNs
⚡
Transformers
💬
LLMs
👁️
Computer Vision
🎯
YOLO
📷
OpenCV

LLM Frameworks

🔗
LangChain
🕸️
LangGraph
🦙
LlamaIndex
🤖
OpenAI API
🧬
Claude
👥
CrewAI
⚙️
AutoGen

Experience Preview

CSIRO's Data61 AI Engineer

Jun 2024 - Jul 2025

Created a multimodal deepfake analysis system integrating CNN/Vision Transformer detection, explainable AI techniques (SHAP, LIME) and image-to-text captioning with CLIP and attention mechanisms. Implemented a large-language-model (LLM) contextual explanation module using Llama 3.1 and Ollama with custom prompt engineering to generate forensic insights.

PyTorch Deep Learning Computer Vision LLMs Llama 3.1

LINQIA AI Engineering Intern

Jan 2025 - Feb 2025

Led the development of algorithms that analysed millions of creator profiles and matched their content styles with brand campaign requirements using advanced natural-language processing and computer-vision techniques. Built an LLM-driven system to generate personalised campaign briefs for creators based on historical performance, brand guidelines and campaign objectives.

NLP Computer Vision LLMs Machine Learning Python

UQ Gas & Energy Transition Centre Data Analyst

Jan 2025 - Feb 2025

Built time-series pipeline to quantify methane-measurement accuracy drift vs temperature, humidity, windspeed; extracted emission patterns, trends and anomalies for sustainability research. Validated 2D/3D anemometer data (sensor agreement, error/variance analysis); refactored/reviewed codebase to improve reliability, reproducibility and analysis performance.

Python Machine Learning Data Analysis Energy

UQ Gas & Energy Transition Centre Software Engineer

Jun 2024 - Jul 2024

Developed a software tool to track and analyse energy-project trends using AEMO data, improving insight into energy transition dynamics. Designed a machine-learning approach for energy transition model calibration that reduced estimation errors by approximately 15% compared with traditional methods.

Python Machine Learning Deep Learning Data Analysis LiDAR

Publications Preview

DF-P2E: Deepfake – Prediction to Explanation

ACM Multimedia 2025

Saakshi Gupta, et al.

Contributed to a multimodal, explainable and interactive deepfake detection framework for non-expert users. The system integrates visual saliency maps, image captioning and narrative explanations to make AI decisions transparent.

View Preprint →

An Advanced Deepfake Detection System leveraging DeepfakeBench and Explainable AI

DFRWS Digital Forensics Conference

Research poster presented at the DFRWS Digital Forensics Conference on advanced deepfake detection methodologies.

View Certificate →

FinGraphRAG: Building a Portfolio Risk & Market Analysis GraphRAG

Coding Nexus

A comprehensive GraphRAG system for portfolio risk and market analysis, combining vector search with knowledge graph reasoning to provide explainable financial insights.

Read on Medium →

View All Publications →

Projects Preview

PR Orchestrator MCP MCP Server

A tools-only MCP server that enables LLM orchestrators to safely automate GitHub pull request creation workflows with isolated sandbox execution, approval gates, and comprehensive safety constraints.

Python MCP SDK E2B SDK FastMCP GitHub API

FinGraphRAG Portfolio Risk & Market Analysis

A comprehensive GraphRAG system for portfolio risk and market analysis, combining vector search with knowledge graph reasoning to provide explainable financial insights.

GraphRAG Neo4j Qdrant FastAPI React Python

FutureVest Stock Analysis | CSSE6400

A web-based stock analysis platform helping investors value stocks and discover growth opportunities with screening, watchlists, alerts, and technical analysis tools.

React TypeScript Flask AWS Docker

Co-Founder Matching UQ Ventures | DECO3801

A mobile app connecting entrepreneurs with potential co-founders using a multi-factor matching algorithm evaluating skills, interests, goals, and collaboration style.

React Native TypeScript Flask Supabase Docker

DF-P2E: Deepfake Detection CSIRO Research

Created a multimodal deepfake analysis system integrating CNN/Vision Transformer detection with explainable AI techniques and LLM-powered forensic insights.

PyTorch Vision Transformers SHAP LIME Llama 3.1