Hwei Geok N.

🎯 Expert-Vetted, Top 1% on Upwork, 100% Job Success. You know the bottleneck you want gone. I build the ML and AI that removes it: production AI agents, LLM and RAG systems, forecasting, and anomaly detection. Shipped, documented, and yours to keep. Nobody understands your business like you do. My job is to see the problem just as clearly. Then build it the way you imagined, and hand it over running in production. Ready for your team to trust and extend. Clients bring the What. I take care of the How. What clients say: → "Hwei is the proverbial needle in a haystack. She gets it. She gets it done." → "She took our vision as her own and brought it to life." → "When she says she'll handle something, it's done." Projects delivered: • Dutch water consortium: anomaly detection that caught real contamination with zero false alarms. Later grown into a 12.3M-measurement platform their own team runs without a developer. • Italian telecom: LLM document validation that cut compliance review from days to minutes. Zero false approvals against a 99.7% precision target. • Fortune 500 consumer brand: shelf-life forecasting platform. Weeks of manual modeling turned into 30-second predictions. • US women's health startup: the AI intelligence layer behind a production multi-agent chatbot. Live ahead of their fundraise. How we work: • Fixed price. Fixed scope. Fixed deadline. • Before any code, we get clear on the problem, the goal, and what success looks like. • Then two-week sprints. Each one ends with something shipped and a decision on what comes next. • Simple MVPs take four weeks. Complex integrations take six to eight. Ongoing work runs as a monthly retainer with one check-in per sprint. • You always know where things stand. Not sure what to build yet? → I also work directly with founders and executives. One working session to separate what's worth building from what's hype, and what your team would need to run it. 🎓 ML and AI since 2018 · 8 published papers · Fraunhofer SCAI and University of Hamburg Send me your project details. I reply within 24 hours with a clear path forward. ================== Tools and capabilities: • Generative AI and LLMs: large language models, RAG (Retrieval Augmented Generation), AI agents and multi-agent systems, document extraction and validation, LLM benchmarking and model selection, prompt engineering, vector databases, LangChain, Pydantic AI, Claude Agent SDK • Machine learning: predictive modeling, anomaly detection, time series forecasting, predictive maintenance, statistical analysis, deep learning, NLP • Engineering and delivery: Python, FastAPI, Docker, AWS, Azure, Render, data pipelines, Streamlit, R Shiny, Gradio • Model APIs: OpenAI, Anthropic Claude, Google Gemini, Hugging Face • Consulting: AI strategy and executive advisory, AI training for leadership teams, AI implementation, AI agent audits, code audits, knowledge transfer