Software engineer who operates above his level — everywhere, every time.

Started coding at AUEB (Management Science & Technology, GPA 8.11) while running a full-time dev job. Graduated with €70k in client revenue already generated and two national hackathon wins. For 3 years I ran parallel full-time tracks — never one thing at a time, always two.

The pattern:

Hackathons are where I first proved the instinct was always to build, not study. No constraints, no bureaucracy, just a problem and a clock. Give me a weekend and a hard problem and I'm in my element.

Built a dynamic taxi allocation system using real Chicago O'Hare Airport data. ML model (logistic regression) predicting aircraft delays by aircraft type, airline, time of day, and destination — pre-positioning taxis before planes landed.

What happened after winning:

• IBM Chief AI Officer invited the team to present the solution in full
• Got on stage at the IBM Greece & Cyprus quarterly summit
• CEO Marios Maniatis handed me the mic in front of a room full of IBM executives

🔗 The chain this started: UNIAI win → IBM CEO stage → Hellas Direct founders scout me → AI Lab → 7-agent claims pipeline → 2,000+ production claims → Digital Finance Awards 2026. One hackathon. Everything that followed traces back to that room.

Built a smart pill dispenser with Face ID in a basement in one month. Stack: TensorFlow/OpenCV (face recognition) + React/Django (information system).

Won first place. Then we didn't stop — because that's what happens when you actually love building things.

→ 📄 Filed a patent on the device
→ 🎓 Entered Lab 2 Market (Columbia University × NTUA program)
→ 🗣️ Ran 70+ customer discovery interviews across pharma and elder care
→ 🔄 Pivoted from consumer product → Medical Device Type 2a for decentralised clinical trials
→ 🗽 Selected as Columbia University finalist, June 2026

7-agent insurance pipeline. Full production. Real numbers.

Pipeline: security → translation → intent classification → extraction → notes generation → audit

Also shipped on the same platform:

• 💬 WhatsApp RAG assistant — AWS Bedrock + Lambda + DynamoDB. Customer queries, fully automated.
• 🔧 Custom OCR engine — Document AI custom classifier (identifies document type) + fine-tuned extractor per document type → structured output. Handles anything the business receives.
• 🔌 MCP server on top of the claims product — their biggest competitive advantage, now tool-accessible for any LLM integration.

Scouted directly by founders Emilios Markou and Michalis Antoniou after the IBM hackathon win. Built the team, designed the architecture, shipped to production, became product owner.

From a basement hackathon win to a patented Medical Device Type 2a. Three years of building, pivoting, and proving the idea had legs.

• 🏆 PreSfhmmy national hackathon winner — where it all started
• 📄 Patent filed on the smart dispenser device
• 🔬 Medical Device Type 2a classification — decentralised clinical trials (at-home stages I & II)
• 🗣️ 70+ customer discovery interviews — pharma companies, elder care operators
• 🗽 Columbia University finalist — Lab 2 Market (Columbia × NTUA), presenting June 2026

Stack: TensorFlow / OpenCV / React / Django

Maintained "Amplio" — Netcompany's global enterprise library, used firm-wide across international offices.

• Selected as the only junior consultant (out of 50+) by Danish HQ for this role
• Formal evaluation: "Exceeds Expectations"
• Java Spring backend + TypeScript custom UI component library
• Running the Hellas Direct AI Lab in parallel the entire time

Full lifecycle ownership on every project — no handoffs, no specialists to hide behind.

• €70k revenue generated from first products delivered
• 3 full projects in ~1 year alongside completing my degree
• Secure banking app integrating gov.gr authentication
• ESG platform tracking 2,000+ KPIs for a major Greek consumer company, schema optimised for Power BI

Languages

AI / Agentic

Backend & Infra

Cloud

Frontend

• 🤖 Building Jarvis — personal AI inference system (local + cloud)
• 🧠 Building a Second Brain — everything documented, building in public
• 🗽 Presenting at Columbia University (Lab 2 Market finalist, June 2026)