Open to opportunities

Nikos Gyftokostas

@nikosgyftokostas

Message

Assistant Manager and Senior AI Engineer specializing in production GenAI, RAG, and agentic systems.

Greece

Message

What I'm looking for

I’m looking for a role where I can lead production GenAI and agentic system delivery end-to-end—designing secure RAG/LLM solutions, mentoring teams, and enforcing reliability through continuous evaluation and monitoring.

I’m a results-driven AI Engineering leader with 5+ years architecting and deploying enterprise-scale GenerativeAI, Machine Learning, and Agentic systems. I specialize in productionizing LLMs and advanced RAG pipelines to process datasets scaling to billions of records.

At EY (Ernst & Young), I currently work as an Assistant Manager (AI/ML & GenAI Systems), directing multiple parallel GenAI initiatives from technical design through production deployment. I mentor cross-functional engineering teams while enforcing robust software engineering standards using SOLID principles and design patterns.

I’ve engineered secure NLP and conversational analytics platforms using MCP servers, SQL, and MongoDB, enabling natural-language access to structured and unstructured enterprise data. I’ve also built an agentic change impact assessment platform with LangGraph and LangChain for stateful, controlled orchestration and decision support.

My focus is reliability and operational excellence: I established continuous evaluation and monitoring frameworks (including Phoenix and Langfuse) to improve LLM reliability, mitigate hallucinations, and support secure, scalable deployments. Before consulting, I led research as a Research Scientist & Machine Learning Specialist at IESL/FORTH, co-authoring publications and translating modeling work into practical predictive systems.

Experience

Work history, roles, and key accomplishments

Current

Assistant Manager, AI/ML & GenAI

Current

Ernst & Young

Jul 2025 - Present (1 year 1 month)

Architects and directs multiple parallel GenAI initiatives, mentoring cross-functional teams and enforcing SOLID principles and design patterns through production deployment. Built secure conversational analytics using MCP servers, SQL, and MongoDB, and implemented agentic orchestration with LangGraph/LangChain alongside continuous LLM evaluation using Phoenix and Langfuse.

GenAI LLM Architecture Lang Chain LangGraph MCP Servers Natural Language Processing SQL MongoDB Conversational Analytics Stateful Agent Orchestration LLM Observability LLM Evaluation Phoenix Langfuse SOLID Software Design Patterns

Senior Data Analyst & ML Consultant

Ernst & Young

Jul 2023 - Jul 2025 (2 years)

Designed early-stage LLM and advanced NLP pipelines for semantic skill extraction from billions of job postings to support enterprise workforce analytics. Engineered PySpark/Databricks recommendation workflows and helped transition analytics into production-grade AI systems using MLflow and Dockerized serving.

LLM Systems Advanced RAG (General LLM NLP Work)Semantic Skill Extraction Recommendation Systems PySpark Databricks MLOps MLFlow Docker Data Pipelines

Data Analyst & ML Consultant

Ernst & Young

Jan 2022 - Jul 2023 (1 year 6 months)

Built robust data processing pipelines and analytical solutions using PySpark to support scalable, distributed enterprise AI applications, including an energy bill prediction platform. Developed and supported deployment of machine learning models for business-critical classification and prediction tasks in live production environments.

PySpark Data Pipelines Machine Learning Classification Prediction

Research Scientist & ML Specialist

IESL / FORTH

Jan 2019 - Jan 2021 (2 years)

Applied statistical modeling and machine learning to analyze spectroscopic datasets and laser-matter interactions using Python. Led work across three concurrent projects and built predictive models for agricultural and recycling applications, achieving classification accuracies exceeding 90%.

Python Machine Learning Statistical Modeling XGBoost SVM PCA predictive modeling

Education

Degrees, certifications, and relevant coursework

University of Patras

Master of Science, Photonics & Lasers

2018 - 2020

Grade: 8.83/10

Activities and societies: Thesis: Classification of extra virgin olive oils via LIBS technique using ns laser pulses combined with machine learning algorithms.

Completed an MSc in Photonics & Lasers at the University of Patras, focusing on classifying extra virgin olive oils using LIBS with ns laser pulses and machine learning algorithms.