I build applied AI systems that run in production, with 5 months of hands-on ML Engineer Intern experience @ Silverpond.

🏆 First Prize @ Agentic AI Build Week 2026 · ex-ML Engineer Intern @ Silverpond · Final-year Computer Science (Data Science & AI) student @ Swinburne.

My experience might sound pretty techy now, but before this, I also worked for 6 months as a Data Analyst Intern @ SW Education. 😁

I started from a data analytics background, where I learned how to clean messy data, understand business context, communicate insights clearly, and work with non-technical stakeholders.

Now, I bring that same communication and problem-solving mindset into ML engineering, AI engineering, AI development, and data science.

In today’s AI world, technical skills alone are not enough.

AI can already do a lot of the heavy lifting: writing code, generating ideas, analysing data, and speeding up development. But humans still need to understand the problem, communicate with people, question the output, monitor the system, explore better solutions, and turn technology into something useful.

That is where I believe I fit well.

I enjoy working across both sides: the technical side of building AI systems, and the human side of understanding what users, teams, and businesses actually need.

That mix of AI engineering, data thinking, communication, and product mindset is what I bring.

• 🛰️ ex-ML Engineer Intern @ Silverpond
  Built agentic AI and computer vision pipelines for transmission-tower inspection work with Energy Queensland. Developed an LLM-controlled SAM3 segmentation workflow, where the LLM reviewed tower images, selected object prompts and confidence thresholds, and called SAM3 as a segmentation tool to generate annotation masks. The workflow reduced manual labelling effort by 50% to 70%. I also built a ResNet18 conductor-material classifier, achieving 89.4% accuracy and 89.5% weighted F1 on a 104-sample test set, and used Claude Haiku via AWS Bedrock + experimented with locally hosted Gemma4 for data-sovereign LLM access.

  💬 Supervisor feedback and internship highlights are available on my LinkedIn.

• 📊 Data analytics foundation
  Built experience in data cleaning, dashboarding, stakeholder communication, and translating messy real-world data into business decisions. I now apply that background to ML and AI systems, especially when explaining model behaviour, evaluation results, and trade-offs to different audiences.

• 🏆 First Prize - Agentic AI Build Week 2026 | Syncnapse Built Syncnapse, an organisational AI memory system that helps companies preserve decisions, context, and knowledge for both employees and AI agents. My main contribution was designing the memory architecture and processing pipeline, including a fast path for immediate retrieval during live interactions and a deeper asynchronous path for processing, consolidation, and long-term memory creation. The system enables AI agents to retrieve persistent organisational context and use that knowledge across conversations and workflows.

• 👥 Led a 6-person capstone team
  Built a predictive-maintenance digital twin where users can query machine health in natural language and run what-if simulations. Diagnosed a failing time-series model with high recall and low precision, then helped drive the system toward a 92% F1 result.

• ⚡ Top 10/32 teams @ Watt The Hack: Energy & AI
  Built SolarCycle AI in an 8-hour hackathon: a live failure-prediction and recovery-routing product for Victorian solar infrastructure. Combined data science, AI, product thinking, and rapid deployment under time pressure.

• 🎓 Bachelor of Computer Science, Data Science & AI @ Swinburne University of Technology
  Graduating Nov 2026 · GPA 3.56/4 · Swinburne International Excellence Scholarship · IELTS 8.0

• 🌱 Currently going deeper on AI engineering, data science, AWS, MLOps, model serving, system design, and production AI systems. I am building the mix of data, software, and deployment skills needed to turn AI ideas into real products.

• 📩 nathanvu.ds@gmail.com

💡 Each badge links to a project where I actually used it - click through.

Languages

AI / ML & Computer Vision

Agentic AI / LLMs

Backend / MLOps / Cloud

Data Science & Analytics

EDA · temporal feature engineering · time-series forecasting · model validation & cross-validation · evaluation (precision / recall / F1 / mAP / RMSE / R²) · star & snowflake data modelling · DAX / Power Query · KPI dashboards

Frontend