Professional Experience

Harry the Hirer

Melbourne, Victoria, Australia

Senior AI Engineer

December 2025 - Present

Spearheaded AI transformation for Australia’s largest event enterprise, converting manual inventory, logistics and planning workflows into high-fidelity, grounded automation for event planning, document generation and evidence-backed decision support.

Key Contributions:

Enterprise multi-agent RAG: Architected an Azure-based platform orchestrating Planner, Retriever and Verifier agents with HITL review, progressive semantic learning, HNSW retrieval, metadata filtering and verifier-grounded cross-checking.
LLMOps and performance engineering: Optimized latency and token spend through context pruning, sliding-window memory, KV caching, regex patternization, async replanning and dynamic multi-LLM routing via Azure AI Foundry.
Advanced retrieval and grounding: Built hybrid retrieval over internal knowledge sources with metadata-aware filtering, evidence matching and output verification to support auditable, hallucination-resistant responses.
OCR and intelligent document processing: Developed GPU-assisted OCR + LLM ingestion pipelines, agentic templating and structured extraction workflows to convert unstructured operational documents into governed data assets.
MCP and ERP integration: Designed protocol-compliant MCP/ERP integrations so AI agents could safely access stock, inventory and planning data across complex multi-step operational workflows.
Production architecture: Engineered modular FastAPI/Pydantic microservices with Docker deployment, MSAL authentication, retries, logging and tool registry patterns, supporting 99.9% operational uptime.

Mizco

Perth, Western Australia, Australia

AI Team Lead

November 2025 - June 2026

Led end-to-end architecture and production delivery of a Trust-First AI platform that converted unstructured RFQ corpora into governed industrial artifacts for compliance-sensitive and SLA-driven workflows.

Key Contributions:

Translated complex RFQ requirements into a pragmatic AI operating model focused on traceability, confidence gates, deployment accountability and resilient fallback behaviour.
Built dual-path agentic inference using ReAct prompting, constrained JSON outputs, FastAPI orchestration and a Next.js operations cockpit for real-time telemetry and concurrency-aware tool execution.
Implemented IDP/ETL standardization, long-context RAG memory, semantic sanitization, self-correction loops and multi-stage backfill/recovery to prevent false completeness.
Designed probabilistic confidence gates and HITL escalation flows to surface uncertainty, support review workflows and maintain governance expectations in industrial environments.

Teqar

Perth, Western Australia, Australia

AI Development Consultant

May 2025 - June 2026

Advise B2B clients on AI transformation by auditing operating models, mapping workflows and converting technical gap analyses into actionable AI roadmaps across ETL/IDP, cognitive CRM, chatbots, virtual call centers, forecasting and automation architecture.

Consulting Scope:

Scoped predictive analytics, computer vision quality assurance, AI workflow automation and data-pipeline modernization initiatives for client-facing operational use cases.
Converted business requirements into implementation-ready solution architectures with measurable productivity targets, technical trade-offs and staged delivery plans.
Helped bridge executive stakeholders, engineering constraints and AI adoption strategy so proposed systems could move from concept to defensible delivery.

Swinburne University of Technology

Melbourne, Victoria, Australia

Researcher

August 2025 - May 2026

Led a 5-person clinical AI lab building specialist medical agents across diagnostics, pharmacology and triage, coordinated by a reasoning orchestrator for grounded, citation-aware medical decision support.

Research Deliveries:

Fine-tuned medical LLMs on 500K+ curated and synthetic cases using knowledge distillation, LoRA/QLoRA, GRPO-style reasoning adaptation and structured augmentation.
Built agentic RAG over EMR/EHR and PubMed with retrieval guardrails, source grounding, citation support and safety validation to reduce hallucination risk in clinical workflows.
Trained an 8M-document biomedical reranker and integrated it into a multi-agent conversation architecture to improve retrieval precision and answer grounding.
Established reproducible evaluation across MedMCQA, PubMedQA and MedQuAD with semantic similarity audits, biomedical embeddings, early stopping and HPC-backed training pipelines, achieving 86% BERTScore.

Harry the Hirer

Melbourne, Victoria, Australia

ML & Data Science

March 2026 - May 2026

Built on the legacy pedigree platform behind world-champion Black Caviar, reimagining thoroughbred pedigree intelligence as an AI-first mate-selection engine that balances genetic merit, compatibility and inbreeding risk.

Key Contributions:

Hybrid recommendation engine: Engineered a 10-signal ensemble for bipartite mate ranking across matrix factorization, GBM temporal ranking, VAE genetic fingerprints, Bayesian nicking, Dosage Theory and compatibility kernels.
Pedigree graph intelligence: Implemented a Shared Ancestor Compatibility Kernel with inverse generational depth, performance weighting and sex-aware cross-line affinity terms.
Graph-aware embeddings: Developed GAT-inspired 64-dimensional pedigree embeddings and BFS graph expansion for scalable candidate retrieval and hierarchical pedigree rendering.
Model governance: Applied strict temporal training splits to prevent leakage in progeny-performance prediction and integrated LLM-powered XAI explanations for breeder-facing recommendations.

AITC

Vietnam / Remote

Senior AI Engineer

June 2025 - November 2025

Architected and led an industrial AI assistant combining LLMs, NLP and OCR/IDP to transform unstructured PDFs into actionable operational guidance, document intelligence and business-ready knowledge workflows.

Key Contributions:

Owned document ingestion, preprocessing, retrieval, deployment and optimization for enterprise knowledge workflows over complex industrial documents.
Mentored data engineers and scaled production AI capability for strategic partner environments including Nvidia, HPA, IBM, Juniper and Cisco Vietnam.
Designed MCP-style assistant workflows that connected local document intelligence, business rules and tool-based automation into controllable AI assistance.

Hai Au Academy

Binh Duong, Vietnam

AI Software Engineer

August 2025 - October 2025

Delivered a bilingual student-management platform for language-school operations, covering enrollment, scheduling, placement testing, attendance, dashboards and role-based administration.

Key Contributions:

Built bilingual Vietnamese/English workflows for students, staff and administrators, reducing fragmented manual administration across school operations.
Integrated MCP-enabled chatbot workflows and NLP helpers for staff-approved automation, translation support, homework assistance and routine administrative tasks.
Designed the system around RBAC, operational dashboards and structured student lifecycle data so the platform could support multi-role and multi-campus expansion.

Pythera

Vietnam / Remote

AI Engineer

January 2025 - September 2025

Built Clara, a Vietnamese medical LLM/VLM assistant trained on 400K+ doctor-annotated cases and 210K+ X-ray/MRI samples, combining clinical dialogue, imaging context and distilled/synthetic supervision for safer local-language healthcare AI.

Key Achievements:

Led EMR-integrated agentic RAG for patient-context grounding, medication-safety checks and auditable clinical support.
Designed retrieval validation and benchmark pipelines reaching >95% recall and >92% precision on labelled medical datasets.
Helped strengthen clinician trust by grounding outputs in patient records, retrieval evidence and structured evaluation workflows.

CPG Seedcom

Vietnam / Remote

AI/Data Scientist

June 2025 - August 2025

Led a 6-member team building a Vietnamese natural-language-to-SQL analytics assistant across MySQL and Oracle for sales, supply-chain and marketing teams.

Key Contributions:

Built the analytics platform using FastAPI, Gemini, Qwen3, MongoDB vector search, schema-aware retrieval and agentic RAG.
Implemented SQL reasoning, validation, self-QA and outcome-based reinforcement loops to turn business questions into trusted database insights.
Reduced manual analysis effort by up to 80% by enabling explainable database-to-decision workflows across business units.

PopTech

Melbourne, Australia

ML Engineer

May 2025 - July 2025

Led a 4-member team delivering an IoT electricity forecasting and control platform for predictive billing, energy optimization and LLM-assisted operational recommendations.

Key Contributions:

Built ingestion and data-quality pipelines with KNN and LinearRegression imputation, anomaly detection and timestamp fallback handling for IoT device streams.
Integrated Temporal Fusion Transformers, XGBoost and DQN-based reinforcement learning to forecast electricity usage and optimize device-control decisions.
Enabled multimodal user interaction through ASR, text interfaces and voice synthesis, creating an interactive LLMOps layer for energy advisory workflows.

The Auburn Coffee

Hawthorn East, Victoria, Australia

Fullstack Developer

April 2024 - September 2025

Developed the official commuter-café platform with Vue.js, Node.js, AWS and MS SQL, delivering an interactive menu, pre-ordering, QR loyalty, staff dashboard, customer feedback and sales analytics.

Impact:

Digitized ordering, loyalty and staff operations for a local café at Auburn Station, supporting faster service and clearer sales visibility.
Supported a ~30% sales uplift through digitalization and improved customer workflow design.
View the live project here.

Nam Á Bank

Ho Chi Minh City, Vietnam

AI/Data Engineer

December 2024 - May 2025

Built fintech AI/data capabilities across OCR, computer vision, LLM/chatbot workflows and data-driven product research for digital banking, onboarding and remittance expansion.

Key Contributions:

Implemented passport OCR and synthetic data generation to accelerate foreign-customer onboarding and support banking services for non-domestic identification scenarios.
Explored remittance-service opportunities targeting overseas Vietnamese markets and contributed product research for potential Australian-market expansion.
Helped redesign mobile-app concepts by shifting complex workflows to web services and improving accessibility for user-friendly digital banking experiences.

Nam Á Bank

Vietnam / Remote

Software Engineer

April 2023 - June 2024

Engineered banking software components across Vert.x backend services and Java Android modules, improving secure API integration, mobile service access, customer onboarding flows and digital banking feature delivery.

Key Contributions:

Developed backend service components and Android modules supporting customer-facing banking workflows.
Improved API integration reliability between mobile experiences and service-layer banking operations.
Collaborated across product and engineering requirements to deliver maintainable fintech functionality in a regulated environment.

iNet Solution

Ho Chi Minh City, Vietnam

Software / Data Scientist

April 2024 - March 2025

Led an LLM-powered analytics pipeline for sentiment analysis, fact-checking and summarization over large document and article collections.

Key Contributions:

Used Obsei, Transformers and cloud architecture to extract actionable business insights from unstructured text sources.
Designed analysis workflows for sentiment detection, document summarization and evidence-aware business overviews.
Supported clients by transforming article/document noise into structured, reviewable intelligence outputs.

VietDan JSC

Vietnam / Remote

ML Engineer

December 2024 - February 2025

Developed and productionized a real-time water-meter OCR system for government-facing utility workflows, converting field images into structured, reviewable meter readings.

Key Contributions:

Built the application with Next.js, Node.js, Flask, YOLOv5x/OCR and 10,000+ annotated samples for detection and reading extraction.
Optimized deployment across Jetson Nano and AWS EC2 to support edge-friendly inference and cloud-backed processing.
Reduced manual processing workload by approximately 35% through automated image capture, detection and OCR workflows.

MJ Industrial

Laverton North, Victoria, Australia

Android Developer

June 2024 - January 2025

Developed and deployed a Kotlin Android workforce and payroll application with Node.js backend and MariaDB database to streamline staff rostering, shift tracking, payment processing and warehouse administration.

Impact:

Reduced manual data entry and accounting workload by centralizing rostering, payroll and shift records in a mobile-first workflow.
Delivered practical backend integrations for workforce operations in an Australian industrial environment.

VietDan JSC

Vietnam / Remote

Cloud Software Engineer

June 2022 - March 2023

Built AWS-hosted cloud services for IoT and ML-enabled applications, supporting production clients with scalable APIs, telemetry ingestion and deployment automation.

Key Contributions:

Supported API gateways, load-balanced backend services, MLOps integration points and cloud deployment workflows.
Implemented monitoring, telemetry and drift-monitoring hooks to improve observability of IoT/data workloads.
Helped containerize and operate backend services for production-facing software delivery.

Alta Media

Vietnam / Remote

DevOps Cloud Engineer

May 2022 - January 2023

Supported AWS/Azure deployment engineering across Dockerized services, Kubernetes workloads, CI/CD automation, environment hardening, monitoring and release workflows.

Key Contributions:

Improved release repeatability by standardizing Docker-based deployment workflows and environment configuration.
Assisted with Kubernetes workloads, CI/CD pipelines and monitoring practices for cloud-hosted products.
Contributed to infrastructure reliability, deployment documentation and backend service operations.

Tien Thanh Co

Ho Chi Minh City, Vietnam

IoT and Software Engineer

January 2020 - August 2020

Engineered cloud-ready IoT and software services across AWS and Azure, containerizing APIs and device-data workloads with Docker and Kubernetes while supporting production deployment, monitoring and scalable service delivery.

Key Contributions:

Built early-stage IoT service components for device data collection, backend processing and operational monitoring.
Supported containerized deployment workflows using Docker and Kubernetes for cloud-ready application delivery.
Worked across hardware-adjacent software, cloud services and production operations, forming a practical foundation in IoT engineering.