Freelance ML Engineer (Hugging Face / Google Colab) – Short-Term PoC

I’m building a proof-of-concept machine learning application that:Ingests activity logs from treasury systems, ERPs and other financial softwareCompares these logs against standard operating procedures (SOPs)Flags errors, out-of-SOP behaviour, and potentially suspicious activityWe’ll build an initial prototype together in Google Colab using Hugging Face, which I will then continue developing on my own.What you’ll doSet up the Python + Hugging Face + Google Colab environment for the projectBuild a first model/pipeline that can:Ingest and preprocess log dataCompare actions against SOP rules or patternsFlag outliers / suspicious activities (e.g. anomaly-style signals)Help identify and select suitable high-quality datasets for this kind of problem (public, synthetic, or anonymised data)Pair-program with me and explain your approach (I have engineering experience and want to learn the workflow)Leave behind:A clean, documented Colab notebook / repoClear notes on next steps so I can continue soloWhat I’m looking forStrong Python skillsExperience with machine learningHands-on experience with Hugging FaceComfortable working in Google ColabAble to handle messy real-world data (logs, timestamps, user IDs, etc.)Bonus: experience finding and evaluating high-quality datasets for ML projectsMost important:You’re proactive and can take a high-level idea and propose a concrete PoC approach, not just wait for detailed specs.How to applyPlease send me:A short intro (who you are and your current role)1–2 examples of relevant work (GitHub, Colab notebooks, portfolio – especially Hugging Face or log/anomaly-type work)Your availability in the week of 17 November 2025