Skip to content
View khadija1317's full-sized avatar

Block or report khadija1317

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
khadija1317/README.md

Hi, I'm Khadija

Final-year engineering student at ESSAI (Tunis), interested in applied AI, NLP and LLM systems. I care as much about knowing when a system can't be trusted as about making it work.

What I'm interested in

  • Evaluation & reliability: Measuring whether a system actually works, and where it fails. Fine-tuning NLLB for English to Arabic showed me a model can improve in-domain and get worse out-of-domain, and you only see it if you test on both.
  • LLMs & NLP: Applications around language models, retrieval and multilingual text.
  • Cost-aware AI: When is a cheaper model enough? I'm particularly interested in the challenge of deciding when a smaller model is sufficient, and how to evaluate that decision.

About me

  • Looking for a 4–6 month PFE starting between January and March 2027, in applied AI, NLP or data.
  • I try to document evaluation scope, design choices and known limitations in my projects
  • Languages: Arabic (native), English (good), French (intermediate).

Currently learning

  • The mathematics behind machine learning: linear algebra, gradients and probability.
  • Transformers and LLMs from the theoretical side.
  • Linux and SQL.

Technical stack

Programming: Python Machine Learning: scikit-learn, XGBoost, SHAP, PyTorch LLM & NLP: Hugging Face Transformers, RAG (BM25 + dense retrieval + reranking), Groq API, ChromaDB, sacreBLEU Document AI: PaddleOCR, EasyOCR Tools: Git, Jupyter, pytest, Gradio

Projects

Project What it does Stack
tunisian-labor-law-rag French-language assistant over the Tunisian Code du Travail and CNSS FAQ; a verifier declines unsupported answers. Small evaluation (21 gold questions) Python · BM25 · bge-m3 · ChromaDB · Groq
massive-routing Routes between a small and a large LLM for intent classification. 77.3% vs 78.0% for the large model alone on 150 test requests, at lower estimated cost Python · scikit-learn · Groq
receipt-intelligence Receipt photo to structured record, with arithmetic checks that flag records it can't trust. Tested on 500 CORD receipts Python · PaddleOCR · LLM extraction
nlp-en-ar-translation Fine-tuning NLLB-200 for English to Arabic. One run: better in-domain, worse out-of-domain PyTorch · Hugging Face
diabetes-readmission Predicting 30-day readmission on 101,766 hospital encounters; four tuned models compared (best ROC-AUC 0.69) scikit-learn · XGBoost · SHAP

📫 Contact: khadija.trabelsi.003@gmail.com

Pinned Loading

  1. massive-routing massive-routing Public

    Cost-aware routing between a small and a large LLM (gpt-oss-20b → gpt-oss-120b) for intent classification on MASSIVE. On 150 test requests it scored 77.3% vs 78.0% for the large model alone (a tie …

    Jupyter Notebook

  2. nlp-en-ar-translation nlp-en-ar-translation Public

    Does fine-tuning NLLB-200 on noisy OPUS-100 help or hurt English→Arabic translation? Single run: in-domain gains (+2.4 BLEU, +1.3 chrF++), out-of-domain losses on FLORES-200 (−2.1 BLEU, −2.0 chrF++…

    Python

  3. receipt-intelligence receipt-intelligence Public

    Receipt photo to structured record, with arithmetic checks that flag records it can't trust. PaddleOCR + rules, tested on 500 CORD receipts .

    Python

  4. tunisian-labor-law-rag tunisian-labor-law-rag Public

    French-language RAG assistant for the Tunisian Code du Travail and CNSS FAQ: hybrid BM25 + bge-m3 retrieval, RRF fusion, cross-encoder reranking, and a two-tier verifier that declines unsupported a…

    Python

  5. diabetes-readmission diabetes-readmission Public

    Predicting 30-day readmission in 101,766 diabetic inpatient encounters (UCI). Four tuned models compared (best test ROC-AUC 0.69, XGBoost; 0.67 in CV), threshold trade-offs, SHAP.

    Jupyter Notebook