23.09.2026 aktualisiert
verifiziert
Premiumkunde
100 % verfügbarFractional CTO & AI Architect | NLP, AI & Machine Learning Expert | Full Stack Developer
Koblenz, Deutschland
Nur Remote
B.Sc. Applied Artificial Intelligence (ongoing) - Universitat OsnabruckSkills
OAuth 2.0MistralAI AgentsRAGGitHub ActionsLLMOpenAIGeminiFine-TuningLLaMAMLOpsLlamaIndexLangChainSupabaseAPIs
AI & Machine Learning (LLM/SLM Fine-tuning, RAG, Agentic AI)
Expertise in LLM/SLM fine-tuning using SFT and RLHF, retrieval-augmented generation (RAG), and agentic AI frameworks such as LangChain and Transformers. Deployment and scaling of large language models for production use cases including Vedic-astrology AI and psychology-based coaching AI.
NLP & Voice AI Engineering
Development of real-time conversational voice agents with STT/TTS pipelines, WebRTC, LiveKit, SIP, and WebSockets. Building production voice systems integrated with CRM, POS, and payment APIs, achieving sub-1-second response latency across up to 20 concurrent calls.
Deep Learning & Computer Vision
Building computer-vision pipelines using CNNs, LoRA fine-tuning, OpenCV, Diffusers, and PyTorch/CUDA. Delivered production systems for medical imaging, clothing imagery automation, and background synthesis, scaling to 24,000+ users.
Backend & Infrastructure
Design and operation of high-concurrency backend systems using FastAPI, Docker, AWS, GCP, NGINX, PostgreSQL, MySQL, and OAuth2/JWT for real-time ML inference and scalable API services.
Frameworks & GPU Computing
Proficiency with PyTorch, TensorFlow, HuggingFace API, CUDA, and TensorRT for training, optimizing, and deploying deep learning models in GPU-accelerated environments.
Real-Time & Sockets
Implementation of real-time communication systems using WebRTC, WebSockets, asynchronous programming, UDP, TCP, SIP, HTTP, and TLS security for low-latency audio and data streaming.
Data Engineering
Construction of ETL pipelines and data workflows using Pandas, NumPy, XGBoost, Matplotlib, Seaborn, and Tableau for data preparation, analysis, and visualization.
Programming
Expert-level programming in Python, with additional proficiency in JavaScript/TypeScript, C++, R, and SQL for diverse software engineering and AI development tasks.
Engineering Leadership
Leadership of engineering teams of up to 12 engineers, covering hiring, mentoring, technical strategy and roadmap definition, stakeholder management, and Agile delivery.
Erfahrung
Berufserfahrung5 - 10 Jahre
Freelancer seit2019
Sprachen
DeutschverhandlungssicherEnglischverhandlungssicher
Projekthistorie
Shipped 21+ SaaS and 34+ B2B applications across Voice AI, Computer Vision, Agentic AI, Healthcare & Pharma, Real Estate, HR, EdTech, and Aviation. Delivered production systems including medical-imaging computer vision (PyTorch/CUDA, Diffusers, LoRA, RAG), a real-time ATC transcription system (UDP + Whisper) for aviation, a RAG-based EdTech voice assistant, and multi-agent frameworks (Solana DeFi and autonomous token generation). Maintained a consistent 5-star rating across freelance platforms and long-term engagements.
Built production voice agents integrated with CRM systems, local databases, and external APIs for fully automated, natural customer conversations. Optimized end-to-end response latency to under 1 second across up to 20 concurrent calls. Engineered non-blocking audio streaming with asynchronous event loops and WebSockets, coordinating real-time TTS/STT I/O and efficient in-call tool execution.
Architected the AI platform that turns product photos into studio-quality flat-lay imagery, scaling to 24,000+ users. Built computer-vision pipelines using CNNs and LoRA fine-tuning for segmentation, enhancement, and background synthesis. Engineered high-concurrency FastAPI and Docker backends serving real-time ML inference.