Collin Aldaibis

The Hague, Netherlands

Building forensic AI in varying R&D projects, leading the engineering and LLM work in collaboration with the police, the Hansken team, and the biological and chemical data labs.

• Fine-tuned Whisper for Dutch forensic audio (ethnolects, dialects, and low-quality recordings), cutting word error rate from ~40% to ~15% on private forensic datasets and teaching it previously unseen slang.
• Pre-trained a Dutch ModernBERT from scratch on Dutch FineWeb / FineWeb-Edu to close a gap in open-source Dutch encoder models with a modern BERT/RoBERTa successor.
• Built and evaluated the Hansken Copilot RAG application, helping investigators search cases of millions of digital traces, fully on-prem with local embedding models and LLMs, evaluated with RAGAS on hand-curated forensic datasets.
• Benchmarked LLMs (GPT-NL, Mistral, Gemma, Qwen) with SURF & TNO; contributed the first Dutch bias benchmark to EuroEval and built a fine-tuning/eval pipeline for a forensic text-classification use case to assess whether GPT-NL could outperform existing models.
• Established local LLM infrastructure for running and fine-tuning models on sensitive data (llama.cpp, Unsloth, Hugging Face, PyTorch), including sandboxed local coding agents and optimised training runs on NVIDIA H200s.
• Founded and chaired the Python Dev Club and LLM Club to raise engineering and LLM literacy across the guild (modern tooling like Ruff and uv, plus safe local agents), and advised NFI leadership on rolling out LLM technology organisation-wide.

Python · PyTorch · HF datasets, TRL, transformers · Unsloth · llama.cpp · Streamlit

Utrecht, Netherlands

Built the company's core AI product, giving clinicians and patients seamless access to AI-powered ECG diagnostics.

• Developed the diagnostics platform end to end, from model serving to clinician-facing UI.
• Supervised several international AI master’s students on deep learning for ECG research theses.
• Organised company events and team activities.

C# · Python · PyTorch Lightning · React (TypeScript) · Docker · Bazel · Terraform · Azure · MLOps

Amsterdam, Netherlands

Delivered ML and full-stack applications for Schiphol as a consultant, owning the relationship with the customer.

• Built and maintained a Flask ML service, a Laravel backend, and an Angular frontend.
• Worked directly with stakeholders at Schiphol to translate business objectives into delivered software.
• Applied NLP techniques, word embeddings, and LLMs to production use cases.

Python · Flask · Laravel (PHP) · Angular (TypeScript) · GCP · Docker · MySQL · CI/CD