Jack van der Vall
Founder of Opusmatic & AI Engineer with a hands-on foundation in technical work preparation, building automation, and Applied Data Science & AI.
Background
With a hands-on background in technical work preparation, building automation, and smart technology, I understand the operational reality of technical firms from the inside out, from AutoCAD isometric piping drawings and cost calculations to parts lists and ERP workflows.
Combining that domain foundation with Applied Data Science & AI, I worked as an AI Engineer during an internship project, developing autonomous data-mining agents at the Erasmus Centre for Data Analytics at Erasmus University and engineering an AI platform for automated contract risk analysis.
Through Opusmatic, I combine technical domain expertise with resilient, version-controlled software architecture (Python, PostgreSQL, PydanticAI, and n8n) to replace manual administrative toil with reliable systems.
“Where we previously spent many hours per contract on manual verification, we now do it in a matter of minutes and, more importantly: the platform automatically flags deviations without missing anything.”
Ivo Alblas
Operations Office Manager
“He is reliable, enthusiastic, and motivated. Brings us valuable insights and innovates company processes in seamless collaboration — over the last 1.5 years leveraging AI techniques. Jack is an asset and a great sparring partner.”
Ruud De Groot
Financial Controller
“Jack placed strong emphasis on governance and human-in-the-loop review on critical steps to prevent hallucination risks, working very independently from problem scoping to a working platform. [...] Jack was an asset to our team.”
Jos van Dongen
Director, Erasmus Data Collaboratory | House of AI
Practical proof
From student projects and internships to systems running in production. This portfolio brings together AI, data engineering and technical work.
Autocrisp
Research for ECDA · Erasmus University
Autocrisp automates data mining based on the CRISP-DM framework using AI agents. From problem formulation and data extraction to model selection and insight generation at scale. Actively used by the Erasmus Centre for Data Analytics.
RotterMaatje 2.0
Student project · Hogeschool Rotterdam · Pauluskerk and Cegaka
Information chatbot for shelter volunteers, with a triage model and a built-in Chainlit interface. Built with PydanticAI agents and Supabase pgvector memory.
Contract analysis platform for an industrial contractor
Multi-agent system for automated analysis of commercial documents. Flags deviations in delivery terms, payment schedules, and scope creep to mitigate operational and financial risk.
agentic-archive
Collection of n8n architectures for multi-model orchestration and RAG-to-SQL.
file-to-vector
Streamlit interface for document ingestion into Supabase pgvector databases.
resume-job-matcher
Semantic matching between resumes and job profiles using sentence transformers.
air-quality-prediction
PM2.5 prediction model using Random Forest and PCA on Dutch RIVM station datasets.
Talk through your process
Tell us which work takes time or leads to errors. In a 30-minute call, we will look at where automation could help.