I will extract entities and classify text using ai and nlp models
Sr Machine Learning Engineer
Over deze dienst
Got messy text data? I'll turn it into something you can actually use.
I build NLP tools that pull structured info out of raw text names, dates, locations, custom categories and get it into a format that works for you. I've been doing this professionally for over a decade, including building the voice search NLU system at Verizon and a text parser powering a real Chrome extension.
What I do:
- Extract entities (people, places, orgs, dates, or custom types like product names or medical terms) from plain text, PDFs, emails, or scanned docs (OCR included).
- Classify text by sentiment, topic, or custom categories single or multi-label.
- Fine-tune or train a custom model when off-the-shelf tools aren't accurate enough. I'll tell you honestly if you need this or not.
- Clean up messy data spelling errors, inconsistent naming ("SF" vs "San Francisco"), the stuff that breaks most pipelines.
Background:
Master's research on text extraction (published at RANLP, CICLing), then industry work building extraction tools and NLU systems.
Send a data sample before ordering if unsure which package fits happy to take a quick look first.
Programmeertaal:
Python
•
SQL
•
Amazon SageMaker
•
Scala
•
Overige
Frameworks:
Scikit-learn
•
Google ML Kit
•
keras
•
PyTorch
•
Panda
Tools:
Jupyter-notitieboek
•
tensorflow
•
MLflow
•
Amazon SageMaker

