Software Engineer · BSc Computer Science ZHAW · AI & Machine Learning
Software engineer focused on AI, RAG systems and backend development. I study computer science at ZHAW and work in parallel at LogObject AG on software for Swiss courts. My passion is building scalable systems that deliver real value – from database migrations to production-ready RAG pipelines.
«Curiosity and determination lead me to new challenges.»
About
A short intro – what drives me and where I work.
I'm Kajaluxan – a software engineer who enjoys turning complex problems into clean, scalable systems. I currently work at LogObject AG on software running in production in Swiss courts, while writing my bachelor thesis at ZHAW on graph-based RAG systems.
My focus is on AI and retrieval systems (LangChain, Qdrant, Neo4J) and on backend engineering with Java Spring Boot, FastAPI and PostgreSQL. From my time as a trained mechanic I bring a sharp eye for clean construction and attention to detail.
Location
Winterthur, Zurich, Switzerland
Focus
AI · RAG · Backend
Studies
BSc CS @ ZHAW
Languages
DE · EN · TA
Key Achievements
Concrete results from recent projects – from database migrations to AI evaluation.
Engineered a complete database migration (100+ tables, MS SQL → PostgreSQL) – deployed to production across 3 client environments.
Designed an evaluation framework for an AI retrieval system with MRR@5: 0.96 and Recall@5: 0.99.
End-to-end delivery of 3 enterprise deployments with data migration, validation and post-deployment monitoring – zero critical incidents.
Professional Experience
From mechanical apprenticeship via IT support to software engineering at LogObject AG.
LogObject AG · Opfikon
Rehaklinik Bellikon · Bellikon
Kantonsspital Winterthur · Winterthur
Coopers IET AG · Zürich
Karl Keller AG · Region Zürich
Ausbildungszentrum Winterthur · Winterthur
Selected Projects
Selected work on RAG, LLMs and backend engineering.
Zürich
AI-powered document analysis platform for a law firm (300+ lawyers) – RAG-based parsing, chunking and preparation of legal documents using LLMs. Built 0→1 in a cross-functional team.
LogObject AG, Opfikon
Privacy-compliant RAG system for the legal domain (fully on-premise) with hybrid search via Qdrant + PostgreSQL (RRF). End-to-end from research prototype to demo-ready system in collaboration with LogObject AG.
ZHAW School of Engineering, Winterthur
Extension of VersionRAG (Huwiler, Stockinger, Fürst – ZHAW): a RAG system that actively detects, indexes and incorporates document versions during response generation.
Blog
Practical articles on RAG, LLM evaluation, vector databases and backend engineering.
What Retrieval Augmented Generation is, why it works and when to use it – from real-world projects.
A sober comparison of Qdrant and PostgreSQL pgvector – performance, setup, scaling and when each one fits best.
Skills & Technologies
The tools and technologies I work with every day.
Education & Certificates
ZHAW School of Engineering · Winterthur
BMS Winterthur · Winterthur
Berufsschule Bülach · Bülach
Innosuisse Business Concept (Module 2)
STARTUP CAMPUS Switzerland
Get in touch
Working on something interesting in AI, RAG or backend? Drop me a line – I'd love to hear about it.