Machine Learning & Data Science
Predictive modelling, forecasting, anomaly detection, feature engineering and decision-support systems.
- Supervised & unsupervised learning
- Time-series modelling
- Statistical validation
Independent AI & Data Science Consulting
From research
to production.
I design and develop intelligent systems that turn complex data into measurable business value, combining rigorous research with pragmatic software engineering.
Expertise
Applied AI for real-world systems.
End-to-end support from problem framing and data exploration to modelling, validation, deployment and technical integration.
Computer Vision
Vision algorithms for industrial and automotive applications, from classical geometry to deep learning.
- Detection & classification
- 3D vision & geometry
- Sensor and camera systems
Generative AI
LLM-powered applications designed around reliability, retrieval, data governance and practical business use.
- RAG architectures
- LLM workflows & agents
- Knowledge extraction
Predictive Maintenance
Data-driven monitoring and early-warning systems for industrial assets and complex equipment.
- Condition monitoring
- Failure prediction
- Remaining useful life
Algorithmic Trading
Quantitative research, statistical modelling, backtesting and automated strategy development.
- Market data pipelines
- Backtesting frameworks
- Systematic strategy research
Research & Prototyping
Rapid technical exploration when the solution is not obvious and off-the-shelf approaches are not enough.
- Feasibility studies
- Proofs of concept
- Research-to-product transition
Selected work
Problems worth solving.
Representative areas of work. Client-sensitive details are intentionally kept confidential.
01
Industrial AI
Predictive Maintenance
Design of machine-learning pipelines for condition monitoring, degradation analysis and early fault detection using operational and sensor data.
02
Perception
Automotive Computer Vision
Camera geometry, visibility analysis, perception pipelines and sensor-oriented engineering for vehicle systems.
03
Quant Research
Algorithmic Trading Systems
Research and development of quantitative models, statistical signals, backtesting workflows and automated execution.
04
Generative AI
Knowledge & LLM Applications
Architecture and prototyping of AI-assisted workflows for information retrieval, document understanding and domain-specific reasoning.
Approach
Research discipline. Engineering pragmatism.
The goal is not to add AI for decorative purposes. It is to build systems that can be measured, understood, deployed and maintained.
01
Understand
Business objective, constraints, data, risks and measurable success criteria.
02
Explore
Data quality, statistical behaviour, baselines and technical feasibility.
03
Build
Models, algorithms, software components and reproducible experimentation.
04
Validate
Robust evaluation against realistic conditions and meaningful business metrics.
05
Deploy
Integration, monitoring and a clear path from prototype to production.
Selected technologies
Python
PyTorch
scikit-learn
OpenCV
NumPy
Pandas
Spark
SQL
Docker
Git
Cloud
MLOps
C++
Databricks
About
Giorgio Gemignani, PhD
Computer scientist and independent consultant specialising in Machine Learning, Artificial Intelligence, Computer Vision and quantitative systems.
My background spans academic research, software development, entrepreneurship and industrial applications. I work at the intersection of modelling and engineering, with a particular interest in problems where data, algorithms and real-world constraints have to coexist.
AIRCD stands for AI Research, Consulting & Development: a compact description of how I work, from technical exploration to deployable solutions.
PhD in Computer Science
Research & industrial experience
End-to-end technical delivery
Selected publications
- 2016A Robust Approach for the Background Subtraction Based on Multi-Layered Self-Organizing Maps, IEEE Transactions on Image Processing 25(11).
- 2015A Novel Background Subtraction Approach based on Multilayered Self-Organizing Maps, IEEE ICIP — best 10% quality paper.
- 2013A Bayesian Approach to Tracking Learning Detection, ICIAP, pp. 803–812.
- 2010Real-Time Stopped Object Detection by Neural Dual Background Modeling, Euro-Par Workshops, pp. 357–364.
Contact
Let’s talk about your project.
Senior technical support for research, prototyping and AI system development.
Emailgiorgio.gemignani@aircd.tech
Based inItalyAvailable for international projects