Vision-Language Model Experts in Germany

Independent engineering practice focused on custom AI systems, production delivery, and fractional technical leadership.

Selected client engagement: Scalutions

• Role: Serve as fractional CTO and hands-on technical lead, responsible for the architecture and agentic infrastructure behind its managed B2B outbound operation.

• Product: Designed and built OutboundLoop, an agentic SDR operating system for research, qualification, personalized outreach, campaign management, human approvals, measurement, and continuous improvement.

• Scope: Own the full system lifecycle—from business processes and agent behavior to context design, model routing, integrations, evaluation, telemetry, reliability, cost control, and production operations.

• Automatic detection of content on various documents
• Recommendation Engine
• Dynamic Pricing

• Developed an evaluation framework for AI-generated multiview driving videos intended for perception and embodied-AI/VLA-related training workflows.
• Designed automated checks for temporal coherence, cross-camera consistency, semantic correctness, and multiview geometric quality, exposing failure modes relevant to autonomous systems.
• Combined classical computer vision, learned visual representations, and vision-language models to convert complex video artifacts into measurable engineering signals.
• Built repeatable benchmarking and failure-analysis workflows to support model comparison, data-quality decisions, and system-improvement discussions.

• Developed and implemented AI-based perception and multimodal systems for real-world environments
• Built, trained, and evaluated Machine Learning and Deep Learning models using Python, PyTorch, TensorFlow, and OpenCV
• Worked with Vision-Language Models (VLMs), Large Language Models (LLMs), transformer-based architectures, and multimodal AI systems
• Applied LoRA-based fine-tuning techniques and experimented with diffusion models for generative and multimodal AI applications
• Developed multimodal perception pipelines using camera, LiDAR, and sensor data
• Designed end-to-end workflows for data processing, model training, evaluation, benchmarking, and robustness analysis
• Utilized HuggingFace Transformers and modern Deep Learning frameworks for AI experimentation and deployment workflows
• Applied GPU-accelerated computing, CUDA-based processing, ONNX, and TensorRT optimization for efficient inference and large-scale model training
• Collaborated with industry partners including Valeo and REHAU on applied AI and intelligent system projects
• Developed scalable AI architectures and prototype software solutions for automation and perception tasks

• Designed and built a multi-tenant Python/Flask API platform with JWT + API-key authentication, scoped access control, and service-level orchestration as the backbone for AI applications.
• Built a multimodal RAG system with hybrid chunking, dense/sparse embeddings, hybrid retrieval, reranking, and vector search to deliver grounded, high-precision responses across enterprise data.
• Productionized AI workflows with Docker, CI/CD, Redis-backed async job tracking, webhook callbacks, external AI/media service integrations, and runtime health/reliability controls.

• Embedded software development for EMS safety boards in medical technology according to IEC 62304 / ISO 13485
• Technologies: C++, OpenCV, Python, Qt6, JTAG, UART, CMake, IEC 62304, ISO 13485

• Insurance service provider*

Challenge: Business-critical documents were processed through external OCR providers, with ongoing costs, dependency, and data privacy risks for sensitive insurance data.

Implementation:

• Architecture and production implementation of an on-premise OCR solution with full data ownership
• Methods for recognizing document structures as the basis for automated further processing
• ML-, NLP-, and LLM/VLM-based information extraction, especially from invoices and quotations

Success: Replaced external providers: full data ownership, GDPR-compliant processing, and 75% lower recurring OCR costs per year

Used technologies: Python, Docker, Microservices, FastAPI, PyTorch, Torchvision, MongoDB, MySQL

• Developing software tools within the EICACS program (LDACS project) supporting secure avionics communication.
• Built Python-based automation and monitoring services to validate AI components under Trustable AI guidelines.
• Designed CI/CD and test pipelines improving reproducibility and reliability across teams.

• Primary Research:
• Developed a compact (<700M parameters) generative vision-language model for whole slide image (WSI) by refining image tokenisation.
• Established an improved evaluation framework, including a curated question-answering dataset and metric selection.
• In preparation for submission.
• Collaboration:
• Conducting research in digital biomarker discovery in computational pathology (CPath) using AI methods.
• Collaborated on projects with international partners, including the Francis Crick Institute (Molecular biomarker prediction in Clear-cell renal carcinoma), HeCOG Greece (Lynch syndrome identification in colorectal carcinoma and Multimodal survival prediction for Prostate adenocarcinoma) and the National Cancer Center Hospital Japan (HIBIRD).
• The work with the Francis Crick Institute is currently being prepared for submission. The collaborative work in Japan has already been published, and the HeCOG projects are ongoing.
• Consortium:
• Manage inter-institutional collaboration and objectives as the KatherLab representative for the LiSYM Consortium.
• Teaching:
• Conducted online workshop sessions for two years at the Clinicum Digitale, educating physicians and medical students on the fundamentals of AI and Python skills.
• Led a multimodal foundation model workshop at the AI in Cancer Research Summer School in Corfu, organized as part of ESAC.
• Presented a talk on vision-language models at the AI in Medicine Summer School, a collaborative event by EKFZ, GENIAL, the TransformLiver Consortium, and ESAC.

• Evaluation and analysis of camera-based traffic light and sign recognition system on various LLM-based autonomous driving systems (LMDrive, BEVDriver)
• Implemented VLM based traffic notice instruction generation unit for closed-loop autonomous driving system which alerts driver in unforeseen driving incidents
• Developed independent LLM-based local chatbot with Llama, DeepSeek and Qwen including MLflow evaluation framework

• Developed a synthetic data generation framework using 3D simulation (NVIDIA Omniverse) and Generative AI (Stable Diffusion) to model and augment industrial surface defects.
• Trained and evaluated Computer Vision models (YOLO, DETR), achieving 94% detection accuracy on real-world samples and demonstrating successful simulation-to-reality transfer.
• Applied domain adaptation to improve simulation-to-reality transfer, enabling scalable Industrial AI for automated quality inspection and reducing manufacturing downtime.

• Developed an end-to-end AI pipeline for analyzing 2D exploded-view drawings using computer vision and deep learning models.
• Integrated YOLO-based object detection (Bounding Boxes, Post-Processing, Overlap Handling) for accurate part and callout detection.
• Applied the Segment Anything Model (SAM) for fine-grained segmentation and separation of individual components.
• Implemented OCR and feature extraction modules, and compared Vision Language Models (VLM) and traditional computer vision approaches in terms of accuracy, runtime, and scalability.

• Designed and implemented multi-modal entertainment toolchains that combine passenger input, vehicle context, large-language models (text-to-text and speech-to-speech) and image generation models to deliver more interactive and immersive in-car experiences.
• Built and orchestrated tools for LLM-based agents, covering session management, background task execution, dynamic user interactions and persistent application state.
• Investigated multi-agent orchestration frameworks for in-car environments, evaluating communication protocols and architectural strategies for coordinated and reliable agent behavior.

• Architected an agentic, real-time offer orchestration engine where specialized agents (retrieval, pricing/optimization, and policy/guardrails) coordinate to personalise promotions across customer touchpoints using RAG with FAISS over Delta Lake and low-latency Databricks Model Serving. Collaborated with product managers and commercial stakeholders to shape the roadmap and evaluate emerging agent patterns for production.

• Designed an agent-based data quality service that orchestrates schema detection, entity normalization, and validator/exception-handling agents to clean multi-retailer SKU feeds at scale. Wrapped model calls in PySpark UDFs for distributed inference, automated via Databricks Workflows and CI/CD.

• Developed a multimodal, agentic extraction pipeline where vision, parsing, and compliance agents collaborate to derive brand, packaging, and volume from scanned images using Claude 3 Sonnet with Swin Transformer encoders. Orchestrated via Azure Event Hub with outputs persisted to Delta Lake.

• Implemented a GS1 taxonomy classification service built around cooperating agents for inference, drift monitoring, and auto-retraining governance using Falcon 180B (LoRA-tuned) with a batch pipeline on Databricks.

• Created a hybrid agent workflow where a retrieval agent surfaces candidate matches via embeddings and a reasoning/verification agent (Mixtral 8x7B) adjudicates receipt-to-SKU alignment, integrated into a streaming Databricks pipeline.

• Built a multimodal attribute inference pipeline structured as cooperating vision-language, rules/consistency, and compliance agents to fill NutriScore, nutrition fields, and packaging types from names and images using LLaMA 3-8B with CLIP embeddings.

• Developed a GenAI-powered orchestration system that ingests recipes from multiple websites, parses ingredients through structured extraction agents, and dynamically links them to real-time retailer offers via tagging, semantic reasoning, and business-rule agents.