Hannes Krauss
Dubai, United Arab Emirates
Hannes Krauss
Senior AI & Security Engineer — Secure LLM, Agents
Category : Artificial intelligence (AI)
I help mid-market and enterprise clients turn AI/LLM ideas into production systems they can actually trust — combining hands-on Python/LLM engineering with a security-first reflex from three years as Team Lead Threat Intelligence.
What I deliver
LLM integration: RAG pipelines, function calling, agent orchestration with LangChain and direct provider APIs (OpenAI, Anthropic, open models)
Production automation: n8n workflows for agent orchestration, LLM pipelines, system-to-system integrations
ML pilots end-to-end: from problem framing to working prototype (most recent: an acoustic-event classification model for an industrial monitoring use case at Cordes & Graefe, 2025)
Internal AI enablement: workshops for engineering teams (Copilot/Claude workflows, prompt patterns, safe usage)
Solution architecture: AI features integrated cleanly into existing application landscapes (REST/SOAP, hybrid on-prem/cloud, secure-by-design)
Why the hybrid matters
AI projects in regulated environments (finance, industry, public-sector adjacent) fail more often on integration and security than on model quality. I bring both in one person: production engineering depth (Python, .NET, TypeScript, cloud) and operational security experience (penetration testing, threat intel, BAIT/MaRisk-context for banking clients). That means RAG systems built with auth and audit trails, agent workflows with safe tool boundaries, and pipelines that survive a real security review.
Tech stack & tools
Languages: Python (expert), C#/.NET, TypeScript, Java, Bash
AI/LLM: LangChain, OpenAI/Anthropic APIs, RAG, prompt engineering, NLP pipelines, audio/ML classification
Automation & CI/CD: n8n (production), Jenkins (incl. Shared Libraries), GitHub Actions, Azure DevOps
Cloud & infra: Azure (Functions, SQL DB, Data Factory), AWS (EC2, S3, Lambda, RDS), Docker, Kubernetes, Linux (admin)
Data: SQL Server, PostgreSQL, MySQL, ETL pipelines
Background
6+ years engineering across full-stack, data, and architecture roles
3 years Team Lead Threat Intelligence at eyeDsec — security mandates incl. banking clients (DACH, anonymized)
Senior Cyber Security Software Architect at Cordes & Graefe (industrial wholesale, 2025–2026)
BSc Computer Science, Hochschule Coburg (2.0)
Engagement model
Remote-first from Dubai (LLC-FZ, EU B2B reverse-charge ready)
Availability: variable, full-time to part-time depending on project
On-site days possible with full travel coverage
Languages: German (native), English (fluent)
Happy to start with a short scoping call — usually 30 minutes is enough to tell whether your project actually needs an LLM, a classical ML approach, or just better automation.
What I deliver
LLM integration: RAG pipelines, function calling, agent orchestration with LangChain and direct provider APIs (OpenAI, Anthropic, open models)
Production automation: n8n workflows for agent orchestration, LLM pipelines, system-to-system integrations
ML pilots end-to-end: from problem framing to working prototype (most recent: an acoustic-event classification model for an industrial monitoring use case at Cordes & Graefe, 2025)
Internal AI enablement: workshops for engineering teams (Copilot/Claude workflows, prompt patterns, safe usage)
Solution architecture: AI features integrated cleanly into existing application landscapes (REST/SOAP, hybrid on-prem/cloud, secure-by-design)
Why the hybrid matters
AI projects in regulated environments (finance, industry, public-sector adjacent) fail more often on integration and security than on model quality. I bring both in one person: production engineering depth (Python, .NET, TypeScript, cloud) and operational security experience (penetration testing, threat intel, BAIT/MaRisk-context for banking clients). That means RAG systems built with auth and audit trails, agent workflows with safe tool boundaries, and pipelines that survive a real security review.
Tech stack & tools
Languages: Python (expert), C#/.NET, TypeScript, Java, Bash
AI/LLM: LangChain, OpenAI/Anthropic APIs, RAG, prompt engineering, NLP pipelines, audio/ML classification
Automation & CI/CD: n8n (production), Jenkins (incl. Shared Libraries), GitHub Actions, Azure DevOps
Cloud & infra: Azure (Functions, SQL DB, Data Factory), AWS (EC2, S3, Lambda, RDS), Docker, Kubernetes, Linux (admin)
Data: SQL Server, PostgreSQL, MySQL, ETL pipelines
Background
6+ years engineering across full-stack, data, and architecture roles
3 years Team Lead Threat Intelligence at eyeDsec — security mandates incl. banking clients (DACH, anonymized)
Senior Cyber Security Software Architect at Cordes & Graefe (industrial wholesale, 2025–2026)
BSc Computer Science, Hochschule Coburg (2.0)
Engagement model
Remote-first from Dubai (LLC-FZ, EU B2B reverse-charge ready)
Availability: variable, full-time to part-time depending on project
On-site days possible with full travel coverage
Languages: German (native), English (fluent)
Happy to start with a short scoping call — usually 30 minutes is enough to tell whether your project actually needs an LLM, a classical ML approach, or just better automation.
Working hours
- Monday:08h00 To 22h00
- Tuesday:08h00 To 22h00
- Wednesday:08h00 To 22h00
- Thursday:08h00 To 22h00
- Friday:08h00 To 22h00
- Saturday:08h00 To 22h00
- Sunday:Not available
![I will build a full-stack web feature or MVP in 1 to 3 weeks]()
From spec to deployed feature — a senior full-stack engineer on your project for one to three weeks. Stacks I work in: frontend in Vue, React, or server-rendered (HTMX-style); backend in D ...![I will build a custom Python automation or data pipeline in one week]()
One week of senior Python engineering to ship the automation or pipeline you've been postponing. Typical scopes: an ETL or data pipeline (Pandas, SQLAlchemy, Airflow if needed), an n8n + Python ...![I will run a focused security review of your web app or API]()
A pragmatic security review for SaaS or fintech teams that need an outside expert eye — not a 200-page generic audit that ends up unread. Scope: external review of one web app or API (authentication, ...![I will integrate an LLM or AI agent feature into your product in 10 working days]()
Ten working days, end-to-end, from "we want AI in our product" to a working production feature you can put in front of users. Typical scope: a RAG (retrieval-augmented generation) flow over your ...![I will audit your codebase and deliver a senior architect's technical roadmap]()
You get an outside senior architect on your codebase for one focused week. I read what's actually there (not just the README), talk to your engineers, and deliver a written technical roadmap ...
At Krasper Technologies (my own freelance practice, since 2026), I work as Founder & Senior Consultant for AI, Cybersecurity and Software Engineering — with the bulk of current mandates centered on AI delivery. I integrate LLMs into production systems (RAG pipelines, function calling, agent orchestration with LangChain and direct provider APIs), build n8n workflows for agent orchestration and system-to-system integrations, and translate AI/LLM ideas into solutions that hold up under a real security review. Alongside the AI track, I deliver Cybersecurity consulting, REST-API architecture and cloud-native integrations on Azure and AWS, and handle the technical product side of client roadmaps — stakeholder alignment, requirements engineering, agile delivery. The hybrid angle is deliberate: production engineering depth combined with the operational security reflex I built up across three years as Team Lead Threat Intelligence.
At Cordes & Graefe (industrial wholesale, 2025–2026), I held a hybrid role as Senior Cyber Security Software Architect & Data Engineer — and increasingly owned the company's AI enablement track. I designed and ran internal AI workshops for the engineering team, covering Copilot/Claude workflows, prompt patterns, and safe LLM usage in business contexts. On the build side, I drove ML pilot projects end-to-end — most notably an acoustic-event classification model that turned a previously manual monitoring task into an automated signal pipeline. Outside the AI track I was responsible for backend development in Python and C#/.NET, Azure cloud architecture (Functions, SQL Database, Data Factory), SQL Server modeling and performance tuning, REST/SOAP integrations to ERP-adjacent systems, and CI/CD workflows — plus technical project planning and stakeholder coordination.
At eyeDsec Cyber Security GmbH (boutique cybersecurity firm, 2022–2025), I worked as Team Lead Threat Intelligence & Cyber Security Specialist — three years building and operating the company's Threat Intelligence practice from the ground up. I led penetration testing and threat-intel mandates for clients in regulated industries, with a strong focus on banking customers in the DACH region (anonymized) — including exposure to BAIT/MaRisk-relevant requirements within the pentest scope. My day-to-day combined hands-on security delivery (vulnerability analysis, pentest execution, security solution design) with team leadership for a small specialist crew, client acquisition, and partner management. That role still shapes how I work today: I treat security as an upstream design concern, not a downstream audit step — an angle that now anchors my AI integration and software architecture mandates.
Academic Foundation: Bachelor of Science (B.Sc.) with a core focus on Software Engineering and Data Engineering. I specialize in transforming complex requirements into structured, maintainable code, leveraging German engineering standards to build robust digital infrastructures.
Please sign in as a customer to give your feedback
Need a service ?
Do you want to hire the services of this Professional ?
Post your need and you will receive dozens of proposals from the Professionals of the community.
