I help companies build and ship AI products—from initial research through production deployment. My work bridges research and production, guiding projects from idea to deployment. This involves everything from initial concept development and proof-of-concept validation to hands-on implementation and cross-functional team collaboration. Driven by a passion for innovation and a commitment to excellence, I deliver AI solutions that create real business impact.
- Agentic AI & Multi-Agent Systems
- LLM Applications & MCP Servers
- RAG Systems & Knowledge Management
- Conversational AI & Chatbots
- Voice Biometrics & Speech Recognition
- B2B Intelligence & Data Products
- AI Strategy & Technical Leadership
PhD in Computer Science, Artificial Intelligence and Robotics
University of Plymouth
BSc in Informatics
Aristotle University of Thessaloniki
I’m a Machine Learning Scientist with over a decade of experience in AI, specializing in Conversational AI, Voice Technologies, and Natural Language Processing.
With a PhD in Computer Science (AI & Robotics) from the University of Plymouth, I bridge the gap between academic research and production systems. My work has resulted in 2 US patents and 7+ publications in top venues including IEEE, Springer, and Applied Sciences.
Currently, I lead R&D at Ada Support, building Generative AI systems that serve millions of customer interactions.
Core Expertise
- Conversational AI — LLMs, chatbots, RAG systems, dialogue management
- Voice & Speech — Voice biometrics, ASR, speaker diarization
- NLP — Text classification, NER, sentiment analysis
- Neural Architectures — RNNs, LSTMs, Transformers, Echo State Networks
- Robotics — Control systems, human-robot interaction
Systems and methods for generating a chatbot
US Patent US11676044B1 · Arnon Mazza, Christos Melidis · Ada Support Inc.
Priority: 2022-08-16 · Granted: 2023-06-13
Patent for automated chatbot generation technology, enabling creation and deployment of conversational AI systems at scale.
Systems and methods for generating automatic training suggestions
US Patent 20230297887A1 · Armand Silviu Gurgu, Christos Melidis · Ada Support Inc.
Filed: 2022-12-15 · Published: 2023-09-21
Patent for automated training suggestion generation, improving chatbot training efficiency through intelligent recommendation systems.
7+ publications in top venues including IEEE, Springer, and Applied Sciences. Research spans neural network architectures, human-robot interaction, conversational AI, and healthcare applications.
Key Research Themes:
- 🧠 Neural Network Architectures — Echo State Networks, RNNs, adaptive systems
- 🤖 Human-Robot Interaction — Intuitive control interfaces (KURE system)
- 💬 Conversational AI — Chatbot generation and dialogue systems
- 🏥 Healthcare AI — Machine learning for functional disorders
PhD in Computer Science, AI & Robotics
University of Plymouth · 2014–2017
Thesis: “Adaptive Neural Architectures For Intuitive Robot Control”
Part of the CogNovo Doctoral Programme
Thesis available on PEARL
BSc in Informatics
Aristotle University of Thessaloniki · 2008–2013
Specialty in Information Systems, Expert Systems, and Database Management
Thesis: “Reinforcement Learning in Robotics”