Projects — end-to-end machine learning, with the code public

MLOPS
FASTAPI
DOCKER
MONITORING

Credit Risk Platform

ROC-AUC 0.78 · cost-based threshold 0.14

End-to-end MLOps platform for credit-default scoring: reproducible training with a versioned model registry, a FastAPI inference API with input validation, PSI drift monitoring and a Prometheus and Grafana observability stack, all containerised with CI. Code at github.com/delcenjo/credit-risk-platform.

PYTORCH
TRANSFORMERS
NLP

Transformer from scratch

Perplexity 6.89 · ablation study

A GPT-style language model implemented from scratch in PyTorch: multi-head attention, causal masking and residual blocks written by hand, plus a byte-pair tokenizer and an ablation study. Code at github.com/delcenjo/transformer-from-scratch.

NEXT.JS
WEBASSEMBLY
CLOUDFLARE
CI/CD

This very site

1.7 MB worker · 142 pages · in-browser lab

This portfolio is hand-built with Next.js and deployed on Cloudflare Workers, with a model lab that runs entirely in the browser and continuous deployment on every commit. Code at github.com/jmweb-org/jmweb-pag-web.

SCIKIT-LEARN
CALIBRATION
TABULAR

Credit Risk Model

Test ROC-AUC 0.779 · 30,000 clients

Credit-default risk model with rigorous evaluation: a leakage-free pipeline, cross-validated model comparison, calibrated probabilities, a cost-based decision threshold and per-segment error analysis. Code at github.com/delcenjo/credit-risk.

RAG
LLM
FASTAPI
DOCKER

AI Insight Assistant

Router: RAG + SQL agent behind one box

An assistant that combines RAG and a SQL agent: it retrieves from a document corpus and queries a database, served with FastAPI and Streamlit and packaged in Docker. Code at github.com/delcenjo/ai-insight-assistant.

RAG
EMBEDDINGS
LLM

RAG Document Assistant

Recall@4 0.88 · live demo on HF Spaces

A retrieval-augmented generation assistant over a document corpus: indexing, semantic retrieval and answers with source citations. Code at github.com/delcenjo/rag-document-assistant.

LLM
SQL
AGENT

LLM SQL Agent

Read-only, with a trace · live demo

An LLM agent that answers questions over a SQL database: it translates natural language into queries, runs them read-only and summarises the result. Code at github.com/delcenjo/llm-sql-agent.

LLM
EVALUATION
BENCHMARK

LLM Eval Harness

Baseline 0.840 acc · macro-F1 0.853

An evaluation harness that compares an LLM against baseline classifiers on a concrete task: reproducible metrics to tell when the LLM is worth it and when it is not. Code at github.com/delcenjo/llm-eval-harness.

NLP
SUMMARIZATION
TRANSFORMERS

Text Summarizer

700-word chunking · live demo

An abstractive text summariser with chunking for long documents: package, CLI and live demo. Code at github.com/delcenjo/text-summarizer.

SCIKIT-LEARN
TABULAR
CHURN

Customer Churn Prediction

CV ROC-AUC 0.845 · the simple model won

End-to-end ML pipeline to predict telecom customer churn: data preparation, modelling and evaluation. Code at github.com/delcenjo/customer-churn-prediction.