About

Applied AI, data, and operational software for regulated environments.

HSE is an Athens-based engineering company. We build production systems for banks, energy operators, public-safety authorities and government agencies. The work has to clear audit, run 24/7, and be defensible under regulatory scrutiny. That constraint shapes everything we do.

§ 01 / What we build today

Three lines of work, one engineering discipline.

We treat AI, data and operational software as a single delivery problem. The model is only as useful as the pipeline feeding it, the system around it, and the audit trail behind every decision.

Line A

Applied AI

LLM-powered agents, retrieval-augmented copilots, federated learning, and ML pipelines built with evaluation harnesses, guardrails and audit trails. Explainable by construction; defensible in regulated review.

Line B

Data platforms

Real-time and historical data backends, governance frameworks, and the integration glue between operational systems. Designed to survive audit, regulator visits, and ten years of organisational change.

Line C

Operational software

Mission-critical applications for the people who actually run banks, grids and public services. Custom where it matters, integrated with what's already there.

§ 02 / Why regulated environments

Because that's where the consequence is.

A model that hallucinates in a chatbot is embarrassing. A model that hallucinates inside a credit decision, a grid-balancing system, or a law-enforcement workflow is a different kind of problem. Our entire engineering practice is built around the second case.

Audit-ready

Evidence chains, not black boxes

Every inference our systems produce has a traceable provenance: which version of which model, against which data, against which prompt or feature vector. Regulators, internal audit and legal teams can reconstruct any decision after the fact.

Operating constraints we live with

  • GDPR & sector-specific data regimes
  • Banking-regulator review
  • EU AI Act, where it applies
  • Evidentiary standards for LEA work
  • Clinical data governance
Production-first

Built to run, not to demo

We don't ship pilots. The systems we build go into 24/7 operation, get owned by the client team, and outlive whichever leadership cycle commissioned them. The work has to survive its own success.

What that means in practice

  • Operability and observability from day one
  • Handover designed into the engagement
  • Boring, maintainable architectures
  • Tests against real data, not mocks
§ 03 / Proof

Twenty-five years of delivery behind the positioning.

The applied-AI focus is recent in name; the underlying engineering discipline is not. HSE was founded in 1999 as an NTUA spin-off and has been shipping into Greek banking, government and public-sector environments ever since. The track record is what lets us take on AI work where the stakes are real.

Banking

Two decades inside Greek banking

Long-running engagements with major Greek banks: real-time market data, document management, loan and credit-card platforms, and the integration work that holds them together. The flagship SpiderStock platform has streamed Athens Stock Exchange data into a Tier-1 bank's e-banking for over a decade.

Selected banking clients

  • Eurobank
  • National Bank of Greece
  • Greek banks for credit & loan platforms
Government

Management systems for Greek public services

Management information systems and citizen-facing platforms for ministries and agencies, designed to strengthen the flexibility and reach of public services without compromising on governance.

Selected public engagements

  • Hellenic Ministry of Defence (MIS)
  • Ministry of Interior portals
  • Meteorological services platform
  • Ministry of Health systems
EU research

Active in Horizon Europe consortia

HSE has been an EU framework programme delivery partner since FP7, with current Horizon Europe work on blockchain forensics, federated learning at the grid edge, and digital tools for the energy transition. Research that ships into production, not white papers.

Selected active projects

  • CryptoACTION
  • HEDGE-IoT
  • DECODIT
  • WeForming
  • EURMARS, TENACITY
Enterprise

Custom systems for large enterprises

Bespoke applications, integrations and analytics for major Greek private-sector organisations across telecoms, energy and adjacent regulated industries.

Selected private clients

  • Cosmote
  • PPC (Public Power Corporation)
§ 04 / Posture

NTUA roots. SME by choice. Senior team in the room.

HSE is privately held and deliberately small. The people in the room with you are the same people writing the code, designing the architecture and signing the deliverables. The NTUA connection keeps the research bench strong and the hiring pipeline serious; the SME posture keeps decision-making fast.

Have a system that has to clear audit?

Whether you are building an AI capability inside a regulated business, planning an EU consortium, or modernising a core operational system: we are the kind of engineering partner you can show the regulator.

Talk to us