I am a Python developer and machine learning engineer specializing in building custom web data extraction pipelines and training task-specific NLP models. If you need to harvest unstructured web data and process it using lightweight, high-efficiency machine learning models tailored to your exact dataset, I can build that pipeline end-to-end.

I hold a BSc in Computer Science and am an MSc Candidate in Artificial Intelligence & Machine Learning at Imperial College London.

Core Technical Stack & Services

• Custom ML & NLP Fine-Tuning: Building and fine-tuning task-specific models (BERT, RoBERTa, DeBERTa, T5) using PyTorch, Keras, and Hugging Face Transformers. I specialize in modifying model architectures (adding/removing custom heads) for sequence classification, token classification (NER), regression tasks, and structured feature extraction on proprietary datasets.
• Resilient Web Extraction: Developing robust web scrapers (Playwright, Selenium, Scrapy, BeautifulSoup) designed to bypass anti-bot configurations, handle heavy JavaScript renders, and manage complex session states.
• Data Engineering: Cleaning, parsing, and structuring messy web data into optimized relational databases (PostgreSQL/SQLAlchemy) or flat data payloads (JSON/CSV).
• Deployment: Packaging scripts and model inference pipelines into Docker containers for clean, automated deployment.

Why Work With Me?

• Cost-Effective AI: Instead of running massive, expensive generative LLM APIs for simple data processing, I train small, highly optimized models that cost pennies to run and offer complete data privacy.
• Production-Grade Code: I write clean, modular Python with robust error handling and structured logging so pipelines run reliably under load.

Rates & Availability

• Hourly Rate: $30 USD / hour (Offering an introductory rate this week for immediate project availability)
• Fixed-Price Tasks: Starting at $100 per discrete, well-defined deliverable.