Dipsan Bhattarai
AI Engineer & Researcher
Computer Science graduate pursuing Master of Applied Artificial Intelligence at Deakin University. Building AI systems that create meaningful, responsible impact.
LLMs
Fine-tuning & Edge AI
Full Stack
Python, Django, React
ML/DL
PyTorch, LangChain
Cloud
Azure, GCP, DevOps
About Me
I am a Computer Science graduate currently pursuing my Master of Applied Artificial Intelligence from Deakin University. With internship and traineeship experience across software engineering and applied AI research, I bring strong analytical thinking and a collaborative mindset.
My current research focuses on fine-tuning Large Language Models on edge devices and investigating Federated Learning frameworks for privacy-preserving, decentralised model training.
Melbourne, VIC, Australia
7+ Professional Certifications
Professional Experience
Full Stack and AI Engineering Intern
PropApp Trading Pvt. Ltd
- Organising and transforming unstructured real estate customer data collected through websites and digital platforms into structured, usable datasets for analysis and business workflows
- Analysing buyer and customer data using Python to identify patterns across customer preferences, locations, finance requirements, and property interests
- Performing customer segmentation and clustering analysis to group buyers with similar behaviours and requirements, supporting targeted property recommendations and customer nurturing strategies
- Developing AI agents integrated with Slack channels to support automated data collection, information extraction, data management, and internal workflow automation
Research Assistant
Deakin University
- Conducting research on fine-tuning Large Language Models (LLMs) on edge devices for efficient on-device AI inference
- Investigating Federated Learning frameworks for privacy-preserving, decentralised model training
- Applying optimisation strategies to enhance model performance on resource-constrained hardware
Software Engineer Intern
Deakin Applied AI Institute (A2I2)
- Designed enterprise workflow automation pipelines using Microsoft Power Automate
- Developed user-facing interfaces in Microsoft PowerApps for non-technical staff
- Utilised AI-assisted development tools including Claude Code, OpenAI Codex, and Vercel v0
- Engineered and validated REST APIs for Power Automate integration with external services
AI Developer
F1Soft International
- Developed intent classification system for customer service chatbot using RAG
- Improved model recall through systematic evaluation using confusion matrices
- Implemented backend services with Python and Django REST Framework
- Managed CI/CD pipelines via Jenkins, Rancher, and ArgoCD
AI Intern
F1Soft International
- Developed multi-page bank statement parser using PaddleOCR for fraud detection
- Built computer vision models using YOLO for object detection tasks
Featured Projects
Email Tone Polisher
Microsoft PowerApps application that rewrites emails according to configurable tones with admin interface for tone and prompt management backed by SharePoint.
- Configurable tones (formal, professional, assertive)
- Admin interface for prompt management
- Automated backend workflow pipelines
Customer Prioritisation System
Generative AI solution for e-commerce with novel Priority Value formula incorporating order frequency and customer spend with configurable weights.
- Dynamic customer ranking by business value
- K-Means clustering for customer segmentation
- Interactive web application for real-time exploration
SecureBank AI Chatbot
Cloud-native banking chatbot on Google Cloud Platform with contextual accuracy and integrated security measures against adversarial inputs.
- Secure, contextual query handling
- Protection against prompt injection
- Full-stack cloud deployment
Technical Skills
Languages
Frameworks
AI Tools
Cloud & DevOps
Databases
Platforms
Certifications & Training
Education
Master of Applied Artificial Intelligence
Deakin University | Geelong, VIC
Relevant Coursework
Deep Learning, Machine Learning, Engineering AI Solutions, NLP, Human-Aligned AI
Bachelor of Technology - Computer Science
Vellore Institute of Technology | Vellore, India
Relevant Coursework
Data Structures & Algorithms, AI, Software Engineering, OS, DBMS
Professional Memberships
- Deakin University Student Association (DUSA)
- Deakin Data Science Club
- Google Developer Group (GDG) Melbourne
- Nepal Engineering Council (Registered Engineer)
Volunteering
- DUSA Thursday Trolley Volunteer (2025 - Present)
- COVID-19 Awareness Campaign - VIT University (2020)
- SET Exhibition Volunteer - Bhaktapur, Nepal (2019)
- Kathmandu World Debates Event Volunteer (2019)
Let's Connect
I'm always open to discussing new opportunities, AI research collaborations, or just having a chat about technology.