Open to opportunities

Vasilis Βalafas

@vasilisalafas

Lead applied AI engineer building production agentic RAG systems—multi-agent orchestration, retrieval, and LLM evaluation.

Greece

Message

What I'm looking for

I’m looking to lead production GenAI/agentic systems end-to-end—architecture, orchestration, retrieval, and LLM evaluation—while shipping hands-on code in an engineering-focused environment that prioritizes quality gates and safe rollout.

I’m a Senior / Lead Applied AI Engineer focused on getting agentic and GenAI systems safely into production. At Accenture, I led GenAI delivery for a Tier-1 European bank and shipped production code hands-on, taking a multi-agent document-processing platform from prototype to production.

I build end-to-end pipelines—agent orchestration, retrieval, extraction, and LLM evaluation/observability—across AWS, Azure, and Databricks. I designed the KYB entity-verification system for Greek corporate documents (about 92% extraction accuracy across 1,000+ entities), and reduced analyst review from ~5 days to under 4 hours, while also shipping a legislation document-processing pipeline that cut end-to-end time from hours to minutes.

Beyond delivery, I’m rigorous about quality gates and rollout safety, creating human-in-the-loop review gates using Ragas/DeepEval for regulated environments. I also co-founded and CTO’d AI products, delivered an immersive VR training platform for the Hellenic Ministry of Citizen Protection, secured five EU-funded R&D projects, and continue research in AI & Optimization through my PhD.

Experience

Work history, roles, and key accomplishments

Current

Team Lead, Applied AI Engineer

Current

Accenture Greece

Jan 2026 - Present (6 months)

Led a team of 4 engineers building GenAI systems for a Tier-1 European bank, owning architecture of a multi-agent platform and shipping production code. Built a multi-agent KYB entity-verification platform and implemented LLM evaluation and human-in-the-loop review gates for regulated rollout.

Multi Agent Systems Agent orchestration Azure Databricks LLM Evaluation Ragas DeepEval Human In The Loop GuardRails LLM Observability CI CD Testing Code Review

Current

Technical Advisor

Current

Innovation Bee

May 2022 - Present (4 years 2 months)

Provided technical advisory support for Innovation Bee, contributing to end-to-end GenAI/agentic system delivery. Supported engineering processes including CI/CD standards and LLM workflows such as RAG and agentic orchestration.

RAG Orchestration Monitoring CI CD

Current

PhD Researcher, AI & Optimization

Current

University of Western Macedonia

Mar 2021 - Present (5 years 4 months)

Conducting PhD research in AI and Optimization on constraint acquisition using a hybrid passive + active learning framework. Developed optimization and reinforcement-learning methods for cloud resource allocation and applied ML to computer vision and IoT monitoring use cases.

Constraint Acquisition Active Learning Reinforcement Learning OR Tools Pyomo CPMpy Computer Vision And IoT Monitoring

Co-Founder & CTO

Innovation Bee

May 2022 - Jan 2024 (1 year 8 months)

Co-founded and led a full-time AI product venture, building and shipping four AI products end to end with agentic design patterns and AI-assisted development. Directed technical delivery, CI/CD standards, and end-to-end LLM workflows including RAG, agentic orchestration, and monitoring.

RAG Orchestration Monitoring CI CD Computer Vision Geospatial AI

Co-Founder & CTO

Gridustry

Oct 2021 - May 2022 (7 months)

Co-founded a decentralized energy startup and led development of a fault-tolerant green-energy certification platform. Built the platform using AWS Serverless components and Kafka event sourcing.

AWS Lambda DynamoDB Cognito Kafka Event Sourcing Serverless Architecture

Software Engineer

UpcoMinds

Jul 2019 - Nov 2021 (2 years 4 months)

Built and scaled backend features for enikos.gr to support high monthly usage. Led development of a decision support system (ICARUS), designing a resilient microservices architecture with Spring Boot.

Microservices Spring Boot Backend Engineering

Education

Degrees, certifications, and relevant coursework

University of Western Macedonia

PhD, Computer Science (AI & Optimization)

2021 -

Activities and societies: Thesis: Constraint Acquisition for Optimization Problems in Data Centers. Funded by the Hellenic Foundation for Research and Innovation (HFRI).

PhD in Computer Science (AI & Optimization) at the University of Western Macedonia focusing on constraint acquisition for optimization problems in data centers (expected Jun 2026), funded by HFRI.

University of Western Macedonia

Integrated B.Eng./M.Eng., Computer Engineering

2015 - 2021

Activities and societies: Thesis: Machine Learning for Object Detection.

Integrated B.Eng./M.Eng. in Computer Engineering at the University of Western Macedonia (Oct 2015–Mar 2021), with a thesis on machine learning for object detection.