Hire the Best Computer Vision Engineers
in the Czech Republic

Computer Vision / ML Engineer (12+ years) delivering end-to-end AI systems: computer vision, 3D LiDAR perception, image & audio processing, geospatial (GIS) analytics, Edge AI optimization with TensorRT & CUDA, and production Python backends on AWS/GCP/Azure. Whatever the data source - camera, LiDAR, satellite, drone, microphone, or IoT sensor - I cover the full lifecycle: model development → GPU inference optimization → containerized deployment with Docker & Kubernetes. One engineer, no handoff gaps between research and production. WHAT I DO 🔹 Computer Vision & Deep Learning • Object detection, segmentation, classification & tracking: YOLOv8/YOLO11, Detectron2, Mask R-CNN, SAM 2 • Video analytics & multi-object tracking: ByteTrack, DeepSORT, re-identification, people/vehicle counting, ANPR/license plate recognition • OCR & document AI, face recognition, pose estimation, defect detection & visual inspection for manufacturing QC • Vision-language models (CLIP, Grounding DINO, VLM pipelines) for open-vocabulary detection & visual search • PyTorch, TensorFlow, OpenCV, ONNX 🔹 3D Perception & LiDAR • Point-cloud pipelines: 3D object detection, tracking, semantic segmentation & scene understanding (PointPillars, CenterPoint, PointNet++) • LiDAR + RGB / multi-camera sensor fusion, intrinsic & extrinsic calibration, BEV perception for autonomous vehicles & robotics • SLAM, visual odometry, mapping & localization (ROS/ROS2) for AMRs, drones & spatial analytics • 3D reconstruction, photogrammetry, NeRF & Gaussian Splatting • Open3D, PCL • KITTI, nuScenes, Waymo 🔹 Edge AI, CUDA & Inference Optimization • TensorRT: low-latency, high-throughput engines (INT8/FP8 quantization, layer fusion, profiling with Nsight) • DeepStream (GStreamer): real-time multi-camera / multi-sensor streaming analytics on Jetson & dGPU • Triton Inference Server: scalable production serving (TensorRT / PyTorch / ONNX backends, dynamic batching) • LLM/VLM serving: TensorRT-LLM, vLLM for efficient batching & deployment • Custom CUDA kernels for pre/post-processing when the profiler says so 🔹 AI & AI Agent & LLM • AI Agents / Voice Agents: CrewAI, AutoGen, Amazon Polly, Deepgram, Rasa AI • LLM Deployment: On Cloud platforms like RunAPod, AWS, GCP • Multi-agent architectures system & Multi- model RAG pipelines 🔹 Image & Audio Processing • Image enhancement & restoration: denoising, deblurring, super-resolution (Real-ESRGAN), HDR, color correction, background removal • Classical + deep pipelines: filtering, morphology, feature matching, camera calibration & rectification • Audio: speech-to-text (Whisper), speaker diarization, voice activity detection, sound event classification, noise suppression • librosa, torchaudio 🔹 GIS & Geospatial AI • Satellite, aerial & drone imagery analytics: land-cover segmentation, change detection, building/vehicle/crop detection • Photogrammetry & mapping: orthomosaics, DEM/DSM generation, georeferencing & CRS handling • GDAL, rasterio, GeoPandas, QGIS, PostGIS, Google Earth Engine 🔹 IoT & Industrial Integration • MQTT pub/sub messaging for lightweight edge-to-cloud telemetry • OPC UA for secure, interoperable industrial data exchange • Modbus, RTSP/ONVIF camera integration 🔹 Python Backend & Cloud Infrastructure • FastAPI backends with async, WebSocket & gRPC streaming for real-time inference APIs • Docker + Kubernetes deployment, Terraform Infrastructure as Code • AWS / GCP / Azure, CI/CD (GitHub Actions), MLOps (MLflow, model registries), observability with Prometheus & Grafana, secure multi-tenant patterns INDUSTRIES: autonomous vehicles & robotics, smart city & traffic analytics, security & surveillance, manufacturing & industrial inspection, agriculture & drone mapping, retail analytics, healthcare imaging. HOW I WORK: clear milestones, honest feasibility assessment up front, profiled benchmarks (not guesses), documented handover, and code your team can maintain. 📩 Whether you need a vision or audio model built, a LiDAR / point-cloud pipeline, satellite or drone imagery analyzed, inference accelerated on GPUs and edge devices, or a complete production backend around your AI - message me with a short description of your project and I'll get back to you quickly with a concrete plan and honest feasibility assessment

You’re looking for a Computer Vision Engineer who can turn real-world challenges into reliable, production-ready solutions. With 10+ years of experience building systems for retail analytics, sports tracking, and edge deployment, I focus on solving the specific business problem behind the code — not just running models. Hi, I’m Jakub, and I’m here to help you bring your vision to life. 🔎 𝗛𝗼𝘄 𝗜 𝗪𝗼𝗿𝗸 1. In-Depth Discovery I begin every project with a thorough conversation, asking lots of questions to clearly define your goals, constraints, and success criteria. This ensures I build exactly what you need to solve your unique problem effectively. 2. Modular Problem-Solving I break complex tasks into smaller pieces, then test multiple approaches to find the most effective solution. This strategy delivers reliable and efficient results you can trust. 3. Leveraging Advanced & Custom Methods My toolkit includes all major computer vision tasks: Classification, detection, tracking, and segmentation. I combine off-the-shelf models with custom-built methods to achieve top-notch performance. 4. Clear, Frequent Communication I keep you updated at every step, provide realistic timelines, and immediately address any hurdles. Even if you’re not technical, I’ll explain everything in plain language so you always know where your project stands. 5. Always Putting You First My clients consistently give me 5-star ratings and glowing feedback. If your requirements stretch beyond my skill set, I’ll be transparent and let you know right away. 🏆 𝗡𝗼𝘁𝗮𝗯𝗹𝗲 𝗣𝗿𝗼𝗷𝗲𝗰𝘁𝘀 & 𝗔𝗰𝗵𝗶𝗲𝘃𝗲𝗺𝗲𝗻𝘁𝘀 • 𝙋𝙤𝙤𝙡-𝘽𝙞𝙡𝙡𝙞𝙖𝙧𝙙 𝘽𝙖𝙡𝙡 𝙏𝙧𝙖𝙘𝙠𝙞𝙣𝙜 (𝙔𝙤𝙪𝙏𝙪𝙗𝙚 𝘾𝙝𝙖𝙣𝙣𝙚𝙡 @𝙋𝙤𝙤𝙡-𝙞𝙨-𝘼𝙧𝙩) Created a tracking system that draws colored lines for each pool ball’s trajectory in real time. This involves ball detection, tracking, color recognition, smoothing, and anti-aliased rendering. One showcased video reached 18K views, demonstrating a blend of precision and style. • 𝙈𝙖𝙧𝙠𝙚𝙩 𝘼𝙣𝙖𝙡𝙮𝙨𝙞𝙨 & 𝘽𝙪𝙨𝙞𝙣𝙚𝙨𝙨 𝙄𝙣𝙩𝙚𝙡𝙡𝙞𝙜𝙚𝙣𝙘𝙚 Built a commercial product that uses cameras in retail stores, coffee shops, and restaurants to gather foot traffic, demographic data, and in-store behavior analytics. Currently deployed in 10 locations, it offers powerful insights for more informed decision-making. • 𝙊𝙥𝙚𝙣-𝙎𝙤𝙪𝙧𝙘𝙚 𝙎𝙚𝙜𝙢𝙚𝙣𝙩𝙖𝙩𝙞𝙤𝙣 𝙏𝙤𝙤𝙡 (𝙐𝙨𝙞𝙣𝙜 𝙎𝘼𝙈 2.1) Combined Meta’s SAM 2.1 with Ultralytics to develop a user-friendly, privacy-focused segmentation tool. Users can apply positive or negative prompts without relying on expensive or riskier cloud services. • 𝙎𝙥𝙤𝙧𝙩𝙨 𝙑𝙞𝙙𝙚𝙤 𝘼𝙣𝙖𝙡𝙮𝙩𝙞𝙘𝙨 a) Soccer: Detected and tracked players to measure goals, ball possession, and movement heatmaps. b) Squash: Measured approximate movement speed by tracking players during training drills. Both applications rely on person detection and tracking, followed by sport-specific analytics. • 𝙇𝙞𝙘𝙚𝙣𝙨𝙚 𝙋𝙡𝙖𝙩𝙚 𝙍𝙚𝙘𝙤𝙜𝙣𝙞𝙩𝙞𝙤𝙣 𝙤𝙣 𝙍𝙖𝙨𝙥𝙗𝙚𝙧𝙧𝙮 𝙋𝙞 Achieved a stable 15 FPS OCR system in full HD on a Raspberry Pi 4B, illustrating how I optimize advanced models for edge devices. • 𝘼𝙒𝙎-𝘽𝙖𝙨𝙚𝙙 𝘾𝙤𝙢𝙥𝙪𝙩𝙚𝙧 𝙑𝙞𝙨𝙞𝙤𝙣 𝘿𝙚𝙥𝙡𝙤𝙮𝙢𝙚𝙣𝙩𝙨 Proficient with Lambda, EC2, and Rekognition to handle tasks like face matching and real-time action recognition at scale. • 𝘿𝙖𝙩𝙖 𝘼𝙣𝙣𝙤𝙩𝙖𝙩𝙞𝙤𝙣 & 𝘿𝙖𝙩𝙖𝙨𝙚𝙩 𝘾𝙧𝙚𝙖𝙩𝙞𝙤𝙣 Adept at using local (CVAT, Make Sense) and cloud-based (Roboflow, Ultralytics Hub) annotation tools. I’ve built large, high-quality datasets (some exceeding 300K labeled instances) to train robust custom models. 🤝 𝗟𝗲𝘁’𝘀 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗲 Ready to bring cutting-edge computer vision to your business or project? • I’ll clarify your goals with detailed conversations. • I’ll design a modular, data-driven solution that meets your needs. • I’ll keep you informed every step of the way with straightforward updates. Send me a message or invite me to your job to explore your vision together. I look forward to bringing your ideas to life!

PROFILE: Skilled and enthusiastic Machine Learning Engineer with an aptitude for self-learning, eager to leverage my expertise in implementing SoTA solutions to drive project success. Have significant experience in computer vision, NLP, time series analysis, recommender systems and basic machine learning. Also have a background in multidimensional statistical analysis(PCA, factor analysis, MDS, correlation analysis, cluster| discriminative analysis). Have good knowledge of differential equations, optimization and operations research probability theory. SKILLS AND KNOWLEDGE: Computer vision object detection/segmentation(Mask R-CNN, U-Net, YOLOv5, YOLOv6, YOLOv8, SSD, Faster R-CNN, etc), image generation(CLIP, stable-diffusion), face verification/detection, face mesh detection(mediapipe), tracking of the objects(STARK, ReID, BotSORT, Bytetracker, etc), OCR NLP speech-to-text(AWS Transcribe, Deepgram), topic modelling(LDA, Amazon NTM, BERTopic), punctuation and capitalization(SpaCy, nltk, HuggingFace), clustering(Faiss, sklearn, Gensim), text generation(BERT, Roberta, GPT-neo, etc.), NER, etc. Recommendation Systems TF Recommenders, Collaborative and Content-based filtering, text similarity, etc. Machine Learning: EDA, XGBoost, KNN, SVM, XGBoost, Random Forest and Decision Tree, PCA and TSNE, factor analysis, MDS, correlation analysis, cluster| discriminative analysis, time series analysis. TECHNICAL STACK: Cloud Platforms •Google Cloud (GCP): Google BigQuery •AWS S3, AWS EC2 Version Control & Collaboration •Git •BitBucket Programming Languages •Python •MATLAB •C++ ML & DL: •PyTorch: Torchserve •TensorFlow: TensorRT, Keras •Mediapipe •ONNX •Stable Diffusion: Diffusers, Dreambooth •Scrapy NLP •HuggingFace •NLTK •SpaCy •Gensim Operating Systems •Linux Computer Vision & Image Processing •OpenCV •Pillow •FFmpeg •Ultralytics •OCR •Defisheye •CVAT Databases •SQL •PostgreSQL API & Web Development •Flask (REST API) •Postman Data Visualization •Seaborn •Matplotlib •Plotly •Dash •Streamlit Data Analysis & Time Series •Pandas •Numpy •CuPy •SciPy •SymPy Tracking •TensorBoard •Weights & Biases(W & B) •MLflow PM Processes: •Agile, Scrum, Kanban. Keywords: | AI Engineer | ML Engineer | Machine Learning | Large Language Models | LLM Developer | LangChain Developer | LangGraph Expert | RAG Developer | Retrieval-Augmented Generation | OpenAI Developer | Claude AI | Gemini AI | ChatGPT API | GPT-4 | Autonomous Agents | AI Chatbots | AI Copilot | AI Assistant | Toolformer | Function Calling | Agentic Systems | ReAct Agents | Prompt Chaining | Agent Routing | Self-Improving Agents | Embedding Models | Semantic Search | Pinecone | Weaviate | Chroma DB | FAISS | Vector Store | Vector Embeddings | Context Compression | AI Pipeline | Custom LLM | Finetuned Models | Model Evaluation | ML Model Deployment | HuggingFace | Model Distillation | Data Annotation | Active Learning | AI UX | AI Product | AI MVP | AI SaaS | SaaS Platform | AI Integration | Python AI | FastAPI Backend | Flask | Django | Node.js Backend | Express.js | Streamlit Apps | Gradio UI | React Frontend | Next.js App | Tailwind | Vite | TypeScript | UI/UX AI | MLOps | DevOps | Docker | Kubernetes | GCP | AWS | Azure AI | Cloud Functions | Cloud Run | EC2 | S3 | RDS | MongoDB | PostgreSQL | Redis | Supabase | Firebase | SQL | NoSQL | NLP Developer | CV Developer | AI for Healthcare | AI for Education | AI for Finance | AI for Legal | AI for Marketing | AI for HR | Real-Time AI | Chatbots for Customer Support | Custom API Integrations | Webhooks | AI Automations | Digital Workers | Intelligent Automation | AI Platform | Scalable Architecture | API Gateway | Microservices | Serverless | Full Stack AI | B2B SaaS | AI Tooling | AI Prototypes | AI POCs | AI for Startups | Early Stage AI | Data Pipelines | ETL Pipelines | Cloud AI | AI Infrastructure | GitHub Copilot | Cursor IDE | Claude Code | Model Monitoring | Explainable AI | AI Compliance | GDPR AI | Token Usage Optimization | Cost-Efficient AI | Multi-Agent Systems | Voice Agents | Audio AI | Whisper API | YouTube Transcript Analysis | Semantic Parsing | PDF Parsing | Document AI | OCR + NLP | Resume Parsing | CRM Automation | AI Scheduling | AI Forms | AI for E-Commerce | AI Chat for Websites | AI Landing Page | AI Email Writer | AI Content Generator | AI Content Calendar | GenAI for Business | GenAI for Sales | GenAI for Support | Generative AI Development | MVP Development | Rapid Prototyping | Custom Tool Development | AI App Builder | AI Workflow Builder | Internal Tooling | Low-code AI | High-load Platforms | Scalable Web Apps | Key Stats & Achievements 🏆 194+ successful projects completed on Upwork with global clients 💲 $5M+ earned on Upwork through high-impact full-stack and AI solutions ⏱ 55,121+ hours billed, proving long-term client trust and engagement 💼 600+ total projects delivered since 2000 across AI, web, and mobile 📚 100% of developers hold engineering degrees, expert-level