Llama Experts in Germany

• Building, implementing and selling automated Data Science solutions such as Scorecard Factory and Forecast Factory
• Implementation of automated end-to-end cloud processes
• Development of LLM and NLP models
• Creation of interactive reports
• Support for national and international large corporations as well as medium-sized companies in implementing ML projects

• Situation: Increasing demand for privacy-compliant AI solutions for clients in the KRITIS and mid-market sector that need to analyze sensitive media content (audio, video, documents) without sending data to public cloud LLMs.

• Task: Design, deployment, and secure operation of a fully self-hosted AI infrastructure including a custom-built digital management platform for automated media analysis.

• Action: Architected and implemented a multi-tier platform on hardened Proxmox infrastructure with frontend (Nuxt 3, Vue 3, TypeScript, Tailwind 4), backend (Laravel 13, PHP 8.4, Sanctum), data storage (PostgreSQL 16, MongoDB 7), caching/queuing (Redis 7, Laravel Queue), AI workers (Python 3.11, Whisper, DeepFace, Librosa), scheduling (Laravel Scheduler/Cron), and local LLMs (Gemma, DeepSeek, Qwen, Mistral, LLaMA, Phi) via OpenWebUI with segmented network access, API hardening, and audit logging following BSI recommendations.

• Result: Fully GDPR-compliant, on-premises AI platform with zero data leakage to third parties.

• Task: Overall responsibility as an external Head of Cyber Security / CISO-as-a-Service for the design, implementation, and continuous improvement of ISMS according to ISO 27001, BSI IT-Grundschutz, and NIS2.

• Action: Built and managed Cyber Defense Centers (CDC) with SOC operations, integrated SIEM solutions (Splunk, Graylog), established risk-based vulnerability management (Qualys, Nessus, OpenVAS), and conducted regular infrastructure, application, and physical penetration tests.

• Result: Audit-ready ISMS for multiple clients and a 60% reduction in critical vulnerabilities within 90 days.

• Task: Design and execution of NIS2 assessments and operational roll-out plans for KRITIS operators.

• Action: Developed an online assessment tool for automated identification of individual weakness profiles, implemented ISMS optimizations, penetration testing, awareness programs, GRC suite deployment, and delivered C-level presentations.

• Result: Accelerated the consulting process by 50% and successfully prepared multiple clients for NIS2 compliance.

• Task: Incident commander for crisis response, forensics, and business recovery in ransomware attacks and APT campaigns.

• Action: Coordinated with state and federal police (LKA, BKA), performed forensic analysis (OSForensics, Wireshark, Kali Linux), executed disaster recovery and BCM strategies, and developed BTC extortion response strategies.

• Result: 100% recovery rate within defined RTO windows and sustainable post-incident security architectures.

• Action: Planned, built, and operated a hardened multi-VM infrastructure (Proxmox, 15+ VMs) with web and mail servers, Graylog, OPNsense firewalls, CRM/ERP and LLM instances, network segmentation, DDoS mitigation, automated patch management, and backup strategies.

• Result: >99.5% uptime over 20+ years and zero compromises.

• Action: Designed coordinated phishing campaigns with five levels of difficulty, developed e-trainings and webinars in a PDCA cycle, and led red and blue teams.

• Result: Phishing click rate reduced from 35% to under 5% within three campaign cycles.

• TrendReel, production LLM agent and RAG system (Python, LangChain, OpenAI, Groq/Llama 3, Claude, FastAPI, Kubernetes, PostgreSQL).
• Designed and built a production multi-step LLM agent system: a script generation agent with a per-platform psychology database, 7 viral narrative frameworks, and structured quality scoring, switching between Claude and Groq backends in real time based on output metrics.
• Implemented multi-provider LLM routing (Claude primary, Groq/Llama 3 fallback) with priority-chain failover and quality-based provider switching, achieving 95% inference cost reduction while holding measurable quality thresholds.
• Built an advanced RAG-style retrieval pipeline with per-platform knowledge bases, semantic content matching, and structured output evaluation across 7 decision frameworks, directly analogous to multi-tenant context-based reasoning for enterprise document workflows.
• BrainyAI, adaptive AI learning platform (Python, LangChain, Groq Llama 3.3-70B, OpenAI, Next.js, Supabase, Redis).
• Integrated Groq Llama 3.3-70B with education-level-aware prompting, dynamically adjusting vocabulary depth, citation complexity, and reasoning style across four student proficiency tiers.
• Nexus Prime, multi-tenant SaaS platform for marketing and growth automation (25 modules, 99 backend routers, 153 frontend files).
• Built a 25-module, 99-router multi-tenant SaaS platform covering ad remix, affiliates, WhatsApp inbox, email, and cart recovery, serving four subscription tiers from $199 to $1,999 per month with integrated Stripe, Paystack, and Flutterwave billing.
• AI Video Surveillance Platform, multi-tenant edge and cloud computer vision system currently in active client pitch.
• Designed a multi-tenant AI video surveillance platform combining edge YOLO26 inference on NVIDIA Jetson Orin NX boxes with a central GKE cloud layer (Postgres, Pub/Sub, ClickHouse, R2, Keycloak) for event storage, dashboards, alerting, and multi-tenancy.

• Further development and operation of a central backend service for providing vehicle inventory for international Mercedes-Benz online shops
• Further development of a reservation service for vehicles as part of the checkout process
• Design and implementation of a highly scalable end-to-end test architecture with a focus on maintainability, reusability, and a high number of automated test cases
• Development of a multi-layer test infrastructure with a strict separation of test logic and access layers
• Development of an initialization and caching architecture to significantly speed up local and CI/CD-based test runs
• Development of an AI-supported review architecture for automated evaluation and quality assurance of end-to-end tests
• Development of a dashboard for consolidated display of a vehicle context across multiple backend systems, including AI-generated summaries and compact case analyses
• Integration and further development of connections to various reservation systems via Apache Kafka and REST
• Design of new microservices and support with architecture decisions
• Risk analysis and design of microservice migrations
• Technologies: Java, Spring, Spring Boot, Gradle, Maven, Apache Kafka, REST, OpenAPI, Swagger, Github, Confluence, CI/CD, Microservices, AI, LLMs

• Designed and deployed enterprise Retrieval-Augmented Generation (RAG) applications using LangChain, LangGraph, vector databases, embeddings, and open-source LLMs served through vLLM on GCP GPU infrastructure.
• Built agentic AI workflows using LangGraph with planning, reasoning, tool execution, persistent memory, session management, and Human-in-the-Loop approval mechanisms.
• Developed LLM-powered automation systems integrating BigQuery, SQL pipelines, and external advertising APIs including Meta, TikTok, Amazon, Snapchat, Google, and Pinterest, reducing manual operational workflows.
• Architected multi-agent AI systems for enterprise analytics and decision-support workflows, enabling autonomous task execution and intelligent data interactions.
• Implemented retrieval optimization strategies including multi-retriever architectures, semantic search, context optimization, and query improvement techniques, improving response relevance by approximately 40%.
• Engineered structured prompting strategies, function-calling schemas, and validation workflows to improve reliability of multi-step LLM applications.
• Designed scalable AI services using Python, FastAPI, Cloud Run, Pub/Sub, BigQuery, Docker, and cloud-native deployment architectures.

• Converts natural language into JQL using RAG and pgvector. Returns structured JSON with a query, chart spec, and plain-text answer. A two-stage router answers general questions without touching the JQL pipeline at all.
• Interchangeable LLM backends: Ollama, vLLM, Groq, Anthropic Claude, AWS Bedrock - switchable at runtime, no code changes. Self-healing JQL: on Jira validation failure, feeds error back to LLM, retries up to 4 times. OCI Vault for secrets. Deployed on Oracle Cloud A1 with GPU inference over Tailscale private network. Open source.

• Provided technical leadership for a 4-engineer team delivering 3 major client platforms in 12 months with microservices architecture and scalability solutions — 100% of scoped majors shipped ahead of schedule vs. planned milestones (baseline: prior releases often slipped 1–2 sprints).
• Ran AI model evaluation and model outputs evaluation on LLM/AI vendor APIs: safety, completeness, instruction adherence, and groundedness review before go-live; cut escaped bad outputs in AI-integrated release checklists from recurring UAT findings to near-zero on final promote.
• Drove API development and performance optimization for payment, exchange, and AI services; fail-closed error handling and payload validation reduced integration rework cycles by ~35% vs. the first AI integration pass.
• Applied software testing, testing frameworks, code quality assurance, and code refactoring with continuous integration gates; first-pass PR acceptance improved across the team and production hotfixes on AI adapters dropped noticeably after review standards landed.
• Owned DevOps practices: Docker, GitHub Actions, Jenkins-compatible pipelines, and version control workflows — cut deployment time ~50% vs. pre-automation baseline and stabilized releases across 3 client environments.
• Implemented verifier/oracle-style pass-fail checks in container sandboxes (Harbor/Terminal-Bench aligned); wrote technical documentation so failures cleared in one review cycle.
• Led cross-functional collaboration with product and client stakeholders; translated AI evaluation scores and risk findings into plain-language briefs for non-technical partners, unblocking go/no-go decisions without extra engineering meetings.
• Used agile methodologies for sprint planning and backlog ownership; mentored engineers so mid-level contributors owned AI adapter modules independently by mid-engagement.

• ▶ Built real-time voice assistant (STT → LLM → TTS pipeline) benchmarking and evaluating multiple STT providers including faster-whisper and Azure Speech. achieved sub-3s latency, Groq API (Llama 3) with multi-turn memory - directly handling edge cases in dictation, names and passcode recognition.

• Spearheaded the integration of a custom Transformer-based encoder into the AFFGANwriting pipeline, replacing the legacy VGG19 architecture to capture richer, high-fidelity writer-style representations.
• Boosted user-study pick-rates by 40%, demonstrating a significant leap in the perceptual quality and realism of the generated handwriting compared to the baseline model.
• Enhanced OCR performance by 20% by implementing a teacher-student framework that leveraged a TrOCR benchmark model for auxiliary training alignment

• Built and operate a hybrid Kubernetes platform across bare metal and cloud to validate multi-GPU workloads, security-zone isolation, and disaster recovery.
• Operate self-hosted AI coding agents in the platform's Git workflow, from issue triage to pull-request review; every change is gated by manifest diffs and policy checks in CI.
• Co-developed a sensor-fusion and GPU edge-inference platform selected by the European Defense Tech Hub from 50 solutions for field testing.

• Acted as technical consultant to optimize multi-layer ensemble models combining ResNet, CNN-BiGRU-Attention, and XGBoost.
• Guided implementation of a Logistic Regression meta-learner to solve class imbalance problems, achieving 92.86% accuracy and 0.9644 AUC on PTB-XL and Chapman-Shaoxing datasets.

Co-founded an AI-native language learning startup, defining the product vision, AI architecture and technical roadmap. Designed and built the AI and backend stack, including LLM fine-tuning pipelines, custom agentic workflows, and scalable inference infrastructure. First product currently in private beta.

End-to-end responsibility for all products - concept, architecture, development, and production operation as the sole developer; in addition, customer meetings, proposals, and marketing.

Underwriting Copilot - AI assistant for industrial insurance (in production at customer sites)

• Supports underwriters in analyzing industrial insurance submissions - in production use at an industrial insurer.
• Framework-independent RAG architecture with Hybrid Search (BM25 + pgvector) across large, mixed document sets.
• Two-stage evaluation and observability pipeline (code assertions + LLM-as-Judge) that makes answer quality, retrieval accuracy, and citation integrity measurable in a regression-safe way.

Kaiki Menu Analyzer - Data intelligence platform (in production at customer sites)

• Automatically captures and analyzes menu data from around 25,000 German restaurants.
• Scalable 7-container architecture (FastAPI, partitioned PostgreSQL, Redis/RQ) with LLM-supported extraction of structured data from PDF, HTML, and images.
• Full CI/CD pipelines (GitHub Actions), production cloud deployment, interactive dashboards (Dash).

Kaiki GEO Atlas - GEO platform (in production at customer sites)

• Measures brand visibility across five AI engines (ChatGPT, Gemini, Perplexity, Grok, Claude), each augmented with web search, orchestrated as a DAG workflow pipeline (Dispatcher → Sub-workflows → Scoring → Report) with fail isolation.
• 6-container deployment (FastAPI, Celery, Redis, PostgreSQL); LLM cost estimation, PDF audit report, rule-based cross-signal insights (no extra LLM cost).

Data Pipeline & Analytics Platform - competitive analysis in the automotive aftermarket

• Automated data pipeline with gap analysis algorithms and role-based access control; 230+ tests.
• Backend with FastAPI, PostgreSQL, SQLAlchemy.

Product development (actively in progress)

BankingGPT - AI assistant for complaint management in cooperative banking

• Security architecture at the core: no AI draft reaches the customer without human approval - the approval decision is in auditable code, not in the language model (monotonic: the model may escalate, never downgrade).
• Real agentic building blocks, each with its own boundary: the model chooses tools itself through an MCP server (read-only, allowlist, capped, fail-safe); sensitive cases are handed off via an open A2A protocol (JSON-RPC, Agent Card, message/send/tasks/get; client implemented by me) to a separate specialist agent (securities/law), which never lowers the review requirement (pinned by test).
• Evaluation-driven over ten analysis rounds; uncovered a security flaw through independent review and blind tests that nine automated runs had missed.
• Voice AI frontend, responding live: covered cases are answered in the conversation, sensitive ones escalate before generation; response latency < 7 s measured (local GPU STT/TTS).

Stack & production readiness: Python, pydantic-ai, FastAPI/Celery, PostgreSQL/pgvector, FastMCP, fasta2a, Docker; multi-tenant capable (physical vector isolation per tenant), PII encrypted, OWASP-LLM reviewed, 275 tests, CI/CD; vendor-portable (Ollama / EU Cloud Vertex).

After-Sales Assistant - agentic RAG/GraphRAG assistant on public OEM manuals (automotive after-sales)

• Genuinely agentic on LangGraph: ReAct agent with four tools and conversation memory - the model decides on its own whether to use the manual (RAG, Chroma), a knowledge graph (GraphRAG, Neo4j/Cypher - decodes warning lights), or a workshop/booking service.
• Human-in-the-Loop before the irreversible action: before every appointment booking, the graph pauses (interrupt) and gets the driver's explicit confirmation - the same approval-before-action discipline as in BankingGPT, in a different framework.
• Eval as CI gate: a three-part scorecard (RAGAS grounding + deterministic tool-routing accuracy + DeepEval safety: does the answer mention the warning first when there is a critical warning?) blocks the pipeline; provider-agnostic (OpenAI/Azure/Anthropic), FastAPI with token streaming.

Stack: Python, LangChain/LangGraph, Chroma, Neo4j, RAGAS/DeepEval, FastAPI, Docker.

• Developing a professional scheduler for organizations with specific detailed requirements
• Evaluating different existing software solutions
• Creating a list of technical requirements
• Implementing these requirements
• Selected technologies: WebDAV, CalDAV, Rust, Baikal, OAuth, Keycloak