Initializing AI Systems

|

Leading AI Intelligence & Deploying Production AI Solutions

AI Engineer leading the AI Intelligence department at iProject LLC, deploying various production applications that automated critical business processes. Specializing in LLM integration, predictive analytics, and NLP tools with proven business impact.

7+ Production Internal & External AI Apps

2+ Years AI/ML Experience

MSc Applied Mathematics & Statistics

Get In Touch View CV

About Me

Building AI solutions that drive real business impact

Background

With an Integrated MSc in Applied Mathematics & Statistics from NTUA, I've built my career at the intersection of advanced mathematics and cutting-edge AI technology. My thesis on Econometric Methods of Cryptocurrency Volatility Estimation (R Programming) achieved a 9.5/10 grade.

Currently, I lead the AI Intelligence department at iProject LLC, where I established the AI division from the ground up, bringing enterprise AI capabilities to the organization and enabling company-wide AI adoption.

Key Achievements

Deployed 7+ production AI applications with measurable business impact
Established AI governance and best practices across the organization
Architected scalable solutions integrating OpenAI, Anthropic, Google, and open-source LLMs
Previous experience in fraud detection and risk analysis at Kaizen Gaming

Technical Skills

Expertise in AI/ML, Programming, and Data Analytics

AI & Machine Learning

LangGraph & Multi-Agent Systems
TensorFlow
Hugging Face Transformers
Scikit-learn & NumPy
RAG & Vector Databases (FAISS)
Prompt Engineering & Fine-tuning

OpenAI Anthropic Google AI Meta AI Open-Source LLMs

Programming & Tools

Python (Advanced)
R (Statistical Computing)
SQL & NoSQL Databases
FastAPI & Flask
Git & Version Control
REST APIs & WebSockets

Cloud & MLOps

Microsoft Azure (AI Services)
Docker & Containerization
Model Deployment & Serving
CI/CD Pipelines
GPU Infrastructure Management
API Development & Integration

Data & Analytics

Statistical Analysis & Modeling
Time-Series Forecasting
Pandas, NumPy & SciPy
Data Visualization (Plotly, Matplotlib)
Business Intelligence
A/B Testing & Experimentation

Professional Experience

Building AI solutions across industries

AI/ML Engineer - Data Analyst

iProject LLC - Athens, Greece

Apr 2024 - Present

Established AI Department: Built the AI Intelligence division from scratch
Deployed various production AI applications including NLQ business intelligence tools
Developed predictive analytics engines for F&B businesses
Created AI lead generation systems and specialized chatbots
Architected scalable AI solutions integrating multiple LLM providers
Established AI governance and best practices organization-wide

Risk & Fraud Analyst

Kaizen Gaming - Athens, Greece

Feb 2023 - Mar 2024

Conducted advanced risk analysis and fraud pattern detection
Maintained internal risk intelligence systems (Confluence documentation)
Developed patterns to anticipate market trends and fraudulent behavior
Designed fraud prevention controls reducing risk exposure
Investigated complex fraud cases and established behavioral profiling frameworks

Mathematics & Physics Educator

Self-Employed - Athens, Greece

Sep 2017 - Jun 2023

Enhanced communication abilities through explaining advanced concepts to diverse audiences
Developed teaching methodologies for complex mathematical and physical concepts

Featured Projects

Production AI applications driving business results

AI Data Analysis Platform

Natural Language Query (NLQ) business intelligence tool that enables non-technical users to query databases and generate visualizations using plain English.

Key Features:

Natural language to SQL conversion
Automated data visualization
Real-time business insights
Multi-database support

Click to view architecture

Predictive Analytics Engine

Advanced forecasting solution designed for F&B businesses, predicting daily/hourly orders, product demand, and raw material consumption.

Capabilities:

Time-series forecasting (85%+ accuracy)
Demand prediction models
Inventory optimization
Resource planning automation

Click to view architecture

Testing Platform

Unified testing platform that orchestrates 6 specialized testing modules, providing comprehensive quality assurance and automated security analysis.

Testing Modules:

Lighthouse Performance Auditor
OWASP Security Scanner
AI-Powered Automated Tester
SEO Performance Analyzer
GitHub Commit Scanner
W3C/WCAG HTML Validator

Click to view details

Personal Project

Stock Analysis System

Multi-agent AI investment research platform powered by LangGraph and Claude 4.5 Sonnet. Five specialized AI agents analyze stocks in parallel, self-correct through reflection loops and reach consensus through structured debate.

Key Features:

5 Specialized AI Agents (Fundamental, Technical, Risk, ESG, Sentiment)
Agent Self-Correction Loops (Draft → Reflect → Refine)
Multi-Round Debate System with Convergence Tracking
Real-time WebSocket Streaming
Production-ready FastAPI Backend

View on GitLab

Click to view architecture

Personal Project

PenBot

AI Chatbot Penetration Testing Framework powered by LangGraph, Claude Sonnet 4.5, and Model Context Protocol. Multi-agent security testing system with evolutionary learning, deep reasoning, and comprehensive OWASP LLM Top 10 2025 coverage..

Key Features:

10 Specialized Agents (incl. Evolutionary, Token Soup, RAG Poisoning, Tool Exploit)
Deep Agent Pipeline: Subagent Refinement + Think-MCP Reasoning
1,070+ Attack Pattern Templates with Real-time Mutation
Multimodal Support (Vision/Image Attacks) & Tavily OSINT
OWASP LLM Top 10 2025 Coverage (9/10 categories) & Attack Lineage Tracking

View on GitLab

Click to view architecture

Data Ingestion & Preparation

Historical data loaded from MySQL database or CSV files, cleaned, validated, and transformed into time-series format with automated feature extraction

→

Seasonality & Feature Engineering

ACF/FFT analysis detects patterns; system generates lag features, rolling statistics, calendar features, and cyclical encodings (sin/cos transformations)

→

Ensemble ML Forecasting

Parallel training of 8 models (TBATS, STL+ARIMA, ETS, Theta, XGBoost, LightGBM, CatBoost, Prophet) with walk-forward validation and adaptive weighting

→

Multi-Level Distribution

Core daily forecast cascades into hourly breakdown → category distribution → product-level quantities → raw material consumption using statistical ratios and recipe-based calculations

→

Caching & Optimization

Disk-based forecast caching + client-side state management eliminates redundant computations; hash-based cache keys ensure data integrity

→

Full-Stack Deployment

Flask REST API backend + interactive JavaScript/Plotly.js frontend delivering real-time predictions via multi-view dashboard (85+% accuracy)

Tech Stack:

Python scikit-learn statsmodels XGBoost LightGBM CatBoost Prophet Flask MySQL Plotly.js

Application that serves as a unified testing platform. It acts as a dispatcher that orchestrates the below testing modules:

Performance Auditor

Real-time measuring of Core Web Vitals (LCP, FID, CLS), accessibility scores, SEO metrics, and best practices with historical tracking database and visual reports

OWASP Security Scanner

AI-powered security analysis detecting vulnerabilities across OWASP Top 10 (2021), including SSL/TLS audits, security headers validation, form vulnerability testing, and CSRF/XSS detection with Claude-driven remediation guidance

AI-Powered Automated Tester

Intelligent scenario generator using AI to create and execute dynamic test sequences via Playwright, with natural language test case generation and automated screenshot capture at the end of the test

SEO Performance Analyzer

Comprehensive SEO auditing featuring meta tag analysis, structured data validation, mobile-friendliness checks, Core Web Vitals optimization, and actionable recommendations for search engine visibility

GitHub Commit Scanner

Integrated code review module with OAuth GitHub authentication, analyzing repository commits for security vulnerabilities, code quality issues, and best practice violations with Slack notifications

W3C/WCAG HTML Validator

Standards compliance testing via W3C validation API, identifying markup errors, accessibility violations (WCAG), and HTML/CSS syntax issues with detailed diagnostic reports

User Input

Enter stock ticker (e.g., AAPL, TSLA, NVDA) via web interface

→

Data Acquisition

Finnhub (real-time quotes, news) + Marketstack (1-year historical) + 15 technical indicators calculated

→

Parallel Agent Analysis

5 specialized agents analyze simultaneously: Fundamental (P/E, margins), Technical (RSI, MACD), Risk (Beta, VaR), ESG (sustainability), Sentiment (news)

→

Self-Correction Loops

3 agents use Draft → Reflect → Refine pattern for quality assurance; reflection triggers if low confidence, generic rationale, or default scores detected

→

Conflict Detection & Debate

If score variance > 15.0 → Multi-round debate (Initial → Reflection → Refinement) with convergence tracking; exits when agents align

→

Consensus & Recommendation

Confidence-weighted voting calculates final score; outputs BUY (≥60) | HOLD (40-59) | SELL (≤39) with detailed reasoning

Tech Stack:

LangGraph Claude 4.5 Sonnet FastAPI WebSockets Python Finnhub API Marketstack API Pydantic Pandas NumPy

Key Concepts:

Agent Self-Correction: 3 agents use Draft → Reflect → Refine pattern for quality assurance
Multi-Round Debate: Conditional loops with convergence tracking (max 3 rounds)
Refinement Triggers: Quality detection via numbers, structure, domain terms
Confidence-Weighted Consensus: Higher confidence agents have more influence
Real-time Streaming: WebSocket updates for live analysis progress

Reconnaissance & Config

CLI Wizard Setup → Tavily OSINT gathers target context, tech stack, and policies

→

Multi-Agent Coordination

10 specialized agents vote on strategy. Think-MCP validates consensus against campaign goals.

→

Deep Agent Pipeline

Evolutionary Engine: Mutates successful attacks. Subagents: Apply Domain Adaptation, Psychological Triggers & Stealth.

→

Multimodal Execution

REST API or Playwright (Web UI). Supports Image Attacks and session rotation.

→

Detection & Learning

Two-Layer Analysis + Post-Response Learning. Updates shared Attack Memory for all agents.

→

Reporting & Lineage

Consolidated findings, HTML Reports, and Attack Lineage Trees showing evolution paths.

Tech Stack:

LangGraph Claude 4.5 Sonnet Model Context Protocol Playwright Tavily API FastAPI Python WebSockets

Advanced Capabilities:

Evolutionary Engine: Genetic algorithm (Verbalized Sampling) breeds new attacks from successful parents.
Think-MCP Reasoning: Validates agent consensus and critiques draft attacks before execution.
Token Soup Dialects: Syntactic transformation (Kernel Traces, XML, Logs) to bypass natural language filters.
RAG Poisoning: Tests retrieval-augmented systems with indirect injection and document manipulation.
Tool Exploit: Probes agentic AI systems for privilege escalation and sandbox escape vulnerabilities.
Visual Jailbreaks: Multimodal testing using adversarial images and text-in-image extraction.
Attack Lineage: Tracks the "genealogy" of attacks to visualize how successful vectors evolved.
Universal Connectivity: Connects via REST API or headless browser (Playwright) for full coverage.

System Stats:

10 Specialized Agents: Jailbreak, Encoding, Impersonation, InfoDisclosure, OutputSecurity, Compliance, TokenSoup, Evolutionary, RAGPoisoning, ToolExploit
1,070+ Attack Patterns: Comprehensive library covering OWASP LLM Top 10 2025 (9/10 categories).
Deep Agent Architecture: 4-stage refinement pipeline (Draft → Domain → Psycho/Stealth → Validation).
Smart Reconnaissance: Automated OSINT gathering to identify target tech stack and policies.
Attack Graph: UCB1 algorithm for strategic multi-turn attack planning.

Education & Certifications

Academic background and professional certifications

MSc Applied Mathematics & Statistics

National Technical University of Athens (NTUA)

School of Applied Mathematical and Physical Sciences (MSc Integrated Degree)

2022

Thesis: Econometric Methods of Cryptocurrency Volatility Estimation (R Programming)

Grade: 9.5/10

Professional Certifications

Microsoft Certified: Azure AI Fundamentals

Microsoft

AI-900

2025

Microsoft Certified: Azure AI Engineer Associate

Microsoft

AI-102

Expected 2026

AWS Certified Machine Learning Engineer

Amazon AWS

Associate

Expected 2026

TensorFlow 2: Deep Learning & AI

Udemy

Certificate

2025

Python for Data Science and Machine Learning Bootcamp

Udemy

Certificate

2025

Machine Learning in Production

DeepLearning.AI

Certificate

2025

Machine Learning with Python

IBM

Certificate

2025

AI Agents in LangGraph

DeepLearning.AI

Accomplishment

2025

Long-Term Agentic Memory With LangGraph

DeepLearning.AI

Accomplishment

2025

Pretraining LLMs

DeepLearning.AI

Accomplishment

2025

MySQL for Data Analytics and Business Intelligence

Udemy

Certificate

2024

SQL for Data Science

University of California, Davis

Certificate

2022

Certificate of Proficiency in English

University of Michigan

ECPE - C2 Level

2021

Languages & Soft Skills

Languages:

Soft Skills:

Critical Thinking
Problem Solving
Communication
Team Collaboration
Leadership

Interests & Passions

New AI/ML Technologies

Blockchain Technology

Sports

Travelling

Let's Connect

Interested in AI/ML solutions or want to discuss potential collaborations? Feel free to reach out!

Email Me LinkedIn GitLab

|

About Me

Background

Key Achievements

Technical Skills

AI & Machine Learning

Programming & Tools

Cloud & MLOps

Data & Analytics

Professional Experience

AI/ML Engineer - Data Analyst

Risk & Fraud Analyst

Mathematics & Physics Educator

Featured Projects

AI Data Analysis Platform

Predictive Analytics Engine

Testing Platform

Stock Analysis System

PenBot

AI Data Analysis Platform - Architecture Flow

User Query

NLQ Parser (LLM)

SQL Builder (Python)

Fetch from DB

LLM Data Analysis + Auto Chart

Frontend

Predictive Analytics Engine - Architecture Flow

Data Ingestion & Preparation

Seasonality & Feature Engineering

Ensemble ML Forecasting

Multi-Level Distribution

Caching & Optimization

Full-Stack Deployment

Testing Platform - Unified Testing Modules

Performance Auditor

OWASP Security Scanner

AI-Powered Automated Tester

SEO Performance Analyzer

GitHub Commit Scanner

W3C/WCAG HTML Validator

Stock Analysis System - Multi-Agent Architecture

User Input

Data Acquisition

Parallel Agent Analysis

Self-Correction Loops

Conflict Detection & Debate

Consensus & Recommendation

PenBot - Multi-Agent Security Testing Architecture

Reconnaissance & Config

Multi-Agent Coordination

Deep Agent Pipeline

Multimodal Execution

Detection & Learning

Reporting & Lineage

Education & Certifications

MSc Applied Mathematics & Statistics

Professional Certifications

Microsoft Certified: Azure AI Fundamentals

Microsoft Certified: Azure AI Engineer Associate

AWS Certified Machine Learning Engineer

TensorFlow 2: Deep Learning & AI

Python for Data Science and Machine Learning Bootcamp

Machine Learning in Production

Machine Learning with Python

AI Agents in LangGraph

Long-Term Agentic Memory With LangGraph

Pretraining LLMs

MySQL for Data Analytics and Business Intelligence

SQL for Data Science

Certificate of Proficiency in English

Languages & Soft Skills

Interests & Passions

Let's Connect