Research & engineering since 1998
Research-driven software,
engineered for long-term use.
Egothor is a technology project family spanning search, language processing, cryptography, distributed systems, data platforms and security engineering. Experimental ideas become focused, practical software with explicit engineering trade-offs.
Current work
Focused projects with a clear technical purpose.
Contemporary Egothor work spans search, language, cryptography, software assurance and small infrastructure libraries. The common thread is explicit architecture, disciplined runtime behaviour and evidence that can be inspected rather than merely asserted.
EGOTHOR v4
A new generation of Egothor search infrastructure for environments where clearance, classification, integrity, compartments and conflict boundaries are part of the retrieval problem itself.
Radixor
Fast, deterministic multilingual stemming for Java, built around compact patch-command tries and a clean separation between model compilation and runtime lookup.
ZeroEcho
Modular Java cryptography for resilient data workflows, combining classical and post-quantum capabilities with offline operation, CLI tooling and PKI integration.
MethodAtlas
Source-level test inventory and audit-friendly security classification for software teams that need reproducible evidence, CI integration and controlled use of AI.
conflux
A lightweight Java library combining shared processing context with publish/subscribe updates for clean coordination between independent components.
Engineering continuity
A lineage, not a catalogue.
Egothor has repeatedly moved from research questions to working systems. The technologies changed; the preference for explicit models, measurable behaviour and lean runtime paths did not.
Explore the project lineageWorking principles
Serious engineering is visible in the constraints.
Good software should make its model, operating assumptions and quality evidence understandable. These principles shape both new projects and the way older ideas are revisited.
Determinism where it matters
Reproducible builds, stable outputs and explicit ordering reduce uncertainty in systems that must be tested, audited or operated for years.
Runtime simplicity
Push complexity into preparation and validation when that leaves the production path smaller, faster and easier to reason about.
Evidence over assertion
Benchmarks, tests, structured reports and traceable decisions are part of the engineering model, not decoration added at release time.
The person behind the work
Leo Galamboš, Ph.D.
Cybersecurity transformation, research and software engineering, with a technical background spanning search, distributed systems, networking and data platforms.
Professional profile