Help & Supportcareers@gravityer.com
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Machine Learning and Artificial Intelligence Consultant | 15 years building AI systems | Head of AI @ Electi
Mission-critical AI and Machine Learning systems fail for two reasons: flawed architecture or flawed implementation. >>> I correct both. Most AI initiatives bleed capital on protracted "transformation journeys" that deliver ambiguous results. My function is to provide decisive intervention. I audit, remediate, and optimize high-stakes ML systems to eliminate systemic risk and engineer tangible performance gains. My practice is focused on three core disciplines: > Systemic Audit & M&A Due Diligence: Forensic analysis of data pipelines, neural networks, and ML solutions to identify and quantify sources of bias, instability, performance degradation, and financial inefficiency. > Risk Remediation & Robustness Engineering: Targeted "deep neural graph surgery" to resolve identified flaws. Re-architecting systems for scalability, security, and predictable performance under operational stress. > Performance & Scalability Optimization: Re-engineering models and infrastructure (Python, C++, CUDA, TensorRT) to maximize computational efficiency, reduce operational latency, and unlock unrealized ROI from existing data assets. Production-Grade Generative AI & LLM Systems: A prototype is not a product. I architect and audit Generative AI systems for deployment in high-stakes environments where accuracy, security, and audibility are non-negotiable. > LLM & RAG Frameworks: Engineering and stress-testing Retrieval-Augmented Generation systems to minimize hallucination, ensure factual grounding with proprietary data, and validate information provenance. > Agentic System Architecture: Designing autonomous agent workflows for the auditable execution of complex business processes. The focus is on creating systems that are reliable, predictable, and aligned with strategic objectives, not just technically capable. > Security & Containment: Implementing the architectural and operat
Imperial College London
Master of Science (MSc), Advanced Computing
January 1, 2010 – January 1, 2011
University of Patras
Bachelor of Engineering (BEng), Master of Engineering (MEng), Computer Engineering and Informatics
January 1, 2005 – January 1, 2010
Algolux
AI & Deep Learning Engineer
March 1, 2020 – March 1, 2021
Montreal, Quebec, Canada
Electi Consulting
Managing Partner, Head of AI
March 1, 2019 – Present
Limassol, Cyprus · Hybrid
NavInfo (Europe) B.V.
AI & Deep Learning Senior Engineer, Advanced Research Lab
December 1, 2017 – February 1, 2019
Brabantine City Row
yodiwo
Advisor
December 1, 2017 – January 1, 2019
yodiwo
Principal Engineer, Machine Learning
October 1, 2015 – December 1, 2017
Data Synthesis EU
Principal AI, Machine Learning Consultant
January 1, 2015 – February 1, 2019
Internet
yodiwo
Machine Learning / Computer Vision Consultant
January 1, 2015 – October 1, 2015
UTX Technologies LTD, a VERINT Company
Software Engineer
April 1, 2013 – October 1, 2015
Cyprus, Limassol
SOFTimpact Ltd
Software Engineer
March 1, 2012 – February 1, 2013
Limassol
National Guard of Cyprus
First Lieutenant (Reserve)
June 1, 2003 – August 1, 2005
Deep Learning Techniques
December 1, 2024 – Present
A versatile, modern library providing **advanced deep learning layers, initializers, constraints,** and **analysis tools** for Keras/TensorFlow. Whether you're researching new architectures, experimenting with custom constraints, or analyzing your network activations, DL Techniques brings you cutting-edge components to accelerate and enhance your workflows. https://github.com/NikolasMarkou/dl_techniques
tensorflow-onnx
August 1, 2020 – October 1, 2020
tf2onnx converts TensorFlow (tf-1.x or tf-2.x), keras, tensorflow.js and tflite models to ONNX via command line or python api. TensorFlow has many more ops than ONNX and occasionally mapping a model to ONNX creates issues.
Instasize user behaviour predictive analytics
July 1, 2017 – September 1, 2017
Predicting subscription and churning behaviour of the users of the popular app Instasize by Munkee. ● Big Data Analytics ● Predictive Modelling ● Time Series Analysis
Instasize image analytics and understanding
October 1, 2016 – February 1, 2017
● Analyzing user images ● Extracting quality metrics ● Detecting and recognizing faces ● Classification of user content ● Heavy use of image processing and deep learning
yodigram
February 1, 2015 – October 1, 2018
yodigram is yodiwo's cutting edge B2B live reporting planogram solution. We use cutting-edge image processing technologies based on Deep Learning Networks and in-house developed segmentation algorithms for object detection and classification. The algorithms are developed and optimized from our team so as to achieve optimal results for the retail environment. Get more info and quotes ➡ info@yodigram.com
Satoshi Hunt, Provably fair bitcoin based cryptolottery
January 1, 2013 – July 1, 2013
Designed and build a provably fair cryptolottery on the Bitcoin Blockchain
Shipping-KPI project
March 1, 2012 – January 1, 2013
Developed a platform for benchmarking vessels and fleets between companies.
Hand Tracking / Object Tracking
September 1, 2011 – March 1, 2012
Personal project: Developed a Object detection and tracking framework in C#.NET and OpenCV for controlling a PC with hand gestures and a webcam. Used advanced Machine Learning and Computer Vision techniques ● Cascade HAAR Detection ● Kalman Filtering ● Monte Carlo Mode Estimation ● Active Contours ● Bayesian Inference ● Deformable Object Models ● OpenCV ● Emgu.NET ● C#.NET
Deception detection using facial expressions
May 1, 2011 – September 1, 2011
Deception has been linked to a number of observable behaviors and physiological re- actions but cannot be directly measured. Through several cues explored by several cognitive scientists and psychologists I created a framework for detecting deception using Machine Learning Algorithms ● K-Means + Heuristics ● Expectation Maximization ● Hidden Markov Models ● Dictionary Learning
Open Ocean Waves Simulation
November 1, 2009 – July 1, 2010
Real time, realistic Open Sea simulation using a combination of signal processing techniques, differential equations and high performance GPU programming ● CUDA Programming ● High Performance Computing ● Differential Equations ● Signal Processing (FFT, IFFT, Applying kernels in Frequency Space) ● Implementation of multiple noise models ● Computational Fluid Dynamics (CFD)
Real-Time Simulation and Rendering of 3D Fluids
November 1, 2009 – July 1, 2010
Real time, realistic Smoke (Fluids) simulation and rendering, using differential equations (PDEs) and high performance GPU programming. ● CUDA Programming ● High Performance Computing ● Differential Equations (PDEs) ● Computational Fluid Dynamics (CFD)
Distributed Ray Tracing using BOINC Platform
November 1, 2008 – May 1, 2009
Created a fully working C#.Net raytracing framework and using the BOINC platfrom distributed the workload to different platforms for faster rendering. ● BOINC Platform integration ● Workload balancing algorithms ● Raytracing algorithms, photon mapping
Cultural Fit Analysis
The candidate's extensive experience across various industries (autonomous driving, retail, finance, social media) and roles (engineer, principal, advisor, managing partner) demonstrates a high degree of adaptability and a broad perspective. Their involvement in R&D and proprietary technology development, coupled with contributions to open-source projects, indicates a proactive and innovative mindset. The diverse project types (personal, B2B solutions, consulting) suggest a strong drive for continuous learning and application of advanced techniques, which aligns well with a dynamic, research-oriented culture. The candidate's academic background from Imperial College London and University of Patras further supports a strong foundation for a high-performance environment.
Soft Skills & Operational Fit
The candidate's experience as a Managing Partner and Head of AI at Electi Consulting, along with advisory roles, suggests strong leadership, strategic thinking, and client-facing communication skills. Their project descriptions indicate a problem-solving mindset and the ability to work on complex, cutting-edge technologies. The diverse project portfolio also points to adaptability and a broad operational fit.