AI you can
trust in production
I design the evaluation and reliability layer for RAG and LLM systems — and the QA that keeps software correct — so both behave predictably in production, not just in a demo.
RAG Reliability Audits · AI Security Hardening (OWASP LLM Top 10) · QA & test automation · app & pipeline security (SAST/DAST, CI gates) · GeoAI.
Most AI systems don't fail at the model.
They fail because nobody measures them.
Confident answers with no grounding in your data.
A prompt, model, or code change quietly breaks quality.
Retrieval that quietly degrades as data grows.
Fixed scope. Fixed price. Clear outcome.
Three service lines, one way of working: start with an audit, then harden, automate, and maintain. No open-ended retainers — every engagement has a defined deliverable.
AI Reliability
RAG Reliability Audit
A fixed-scope diagnostic of your RAG/LLM system: where it's fragile, why, and what to fix first.
AI Security Hardening
Ship safely — OWASP LLM Top 10 in practice
- Guardrails & output validation
- Data-leakage prevention
- CI gate for unsafe outputs
LLM Evaluation Pipelines
Automate quality — never ship a silent regression
- LLM-as-Judge pipelines
- Faithfulness / relevancy scoring
- Regression gating in CI (GitHub Actions or Azure DevOps)
QA & Test Automation
QA Audit
Find your baseline — a one-time diagnostic of your test coverage: where you're exposed, why it matters, and what to fix first.
1 app / module · recommendations, not implementation
Book the auditAutomation Starter
Get moving — a turnkey delivery project
- Playwright framework setup (config, conventions, CI)
- 20–40 automated tests for critical user journeys
- CI pipeline integration — GitHub Actions or Azure DevOps (PR gating)
- Docs + 2 review sessions + handover workshop
.NET / React / Node · delivery 4–6 weeks
QA Retainer
Keep it healthy — a monthly partnership
- Guaranteed monthly hours (10 / 20 / 40 h)
- Manual testing of new features before release
- Maintenance & extension of automated tests
- Monthly QA report + Slack / Teams access
Min. 3 months · unused hours don't roll over
App & Pipeline Security
Security & Pipeline Audit
Know your exposure — a one-time assessment of your app and CI/CD pipeline: what's unscanned, what's exposed, and what to fix first.
1 app / repo · posture assessment & recommendations, not a manual pentest
Book the auditPipeline Security Hardening
Close the gaps — SAST, SCA & secret scanning wired into your pipeline, not bolted on later.
- SAST + dependency/secret scanning as PR-blocking CI gates
- DAST baseline scan for web-facing services
- SBOM generation + IaC/container scanning
- Docs + handover workshop
GitHub Actions or Azure DevOps · fix-PRs opened against your repo, not just findings
Security Retainer
Stay current — recurring scans and triage, not a managed SOC.
- Recurring SAST/DAST/SCA scans (monthly cadence)
- Dependency & CVE triage with patch recommendations
- CI security-gate maintenance as tooling evolves
- Monthly security report + Slack / Teams access
Min. 3 months · scoped scan & triage — not 24/7 monitoring or an SLA
GeoAI
AI over spatial data, backed by 17 years in geoscience platforms. A niche where both service lines meet — evaluated for reliability and validated against domain reality, not just unit tests.
From "works in a demo" to "behaves predictably"
RAG & LLM evaluation
LLM-as-Judge pipelines; faithfulness / relevancy / hallucination scoring; retrieval metrics; regression gating in CI (GitHub Actions or Azure DevOps).
RAG reliability
Retrieval quality (hybrid search, reranking), failure modes, and fallback strategies that hold up under real load.
AI security hardening
OWASP LLM Top 10: guardrails, output validation, data-leakage prevention, and a CI gate for unsafe outputs.
QA & test automation
Playwright test frameworks, coverage & risk audits, and CI/CD pipeline integration (PR gating) across .NET, React, and Node apps.
App & pipeline security
App + pipeline hardening: SAST/DAST, dependency & secret scanning, CI security gates, SBOM and supply-chain checks — across GitHub Actions or Azure DevOps.
GeoAI
AI over spatial data — PostGIS, QGIS, WMS/WFS — validated against domain reality, not just unit tests.
Outcome: across AI reliability, app QA, security and GeoAI — from "works in a demo" to "behaves predictably under real conditions."
17 years where wrong outputs had real consequences
For 17 years at SLB (Petrel, DELFI) I built and tested complex geoscience platforms. That shaped how I work: I don't test whether a system runs — I validate whether it's correct, against domain reality.
Today I work as an independent contractor, remote across the EU, based in Bratislava — under TGS Consult s.r.o..
Building software or an AI feature and unsure if it's reliable?
Start with an audit. Book a 15-minute discovery call — we'll scope where your reliability gaps are and whether an audit is worth it.
Prefer LinkedIn?
Book a discovery call on LinkedIn →Bratislava, Slovakia · Remote across the EU
The operator of this website and provider of the services above is TGS Consult s.r.o., IČO 54 396 590, DIČ 2121659496, with registered office at Bleduľová 6756/12, 841 08 Bratislava, Slovakia. Behind this project is Vladimír Fejdi, managing director (konateľ) and owner of the company.