Rakebun L.

I build production AI systems — not demos, not prototypes. If you need an AI engineer who can take a product from architecture through deployment, here's what I bring: → AI Agents & LLM Systems: Built end-to-end agentic workflows integrating OpenAI, Gemini, and ElevenLabs with multi-modal data pipelines (text, audio, images) at scale on GCP. Experience with RAG architectures, tool-calling, memory systems, and automation pipelines. → Backend & APIs: FastAPI and Flask in production. Supabase/PostgreSQL schema design that reduced query latency by 40%. Deployment via GCP with CI/CD from GitHub. → ML Engineering: 4+ years delivering production models — cancer segmentation at 98% accuracy, real-time flood monitoring, multilingual speech-to-text across 4 languages, anomaly detection. PyTorch, TensorFlow, HuggingFace, scikit-learn. → Document Intelligence: RAG pipelines, automated document summarisation, custom GPT systems for domain-specific terminology (70% reduction in processing time for one client). I've maintained a 100% Job Success Score across 4+ years as a Top Rated freelancer. I work best with clients who have a real product problem and need an engineer who can own the AI layer end-to-end. Available 20+ hrs/week. Sydney, Australia (AEST)