Nikita C.

Hardware works on the bench but fails in the field? I make cameras, firmware, apps and servers work as one product: YOLO on NVIDIA Jetson, ESP32 firmware, BLE apps, OTA updates. Clients bring me in when a camera has to understand what it sees, when firmware, app and server were built by different people and don't fit together, or when the previous developer left halfway. $100K+ earned on Upwork · 7+ years building production systems with .NET, C# and Python. I work with my team at DevsX, so electronics, QA and design are covered when the project needs them. 👁️ COMPUTER VISION & EDGE AI Detection and tracking that runs on the device, next to the camera, without cloud delays. - Real-time object detection and tracking with YOLO and OpenCV on NVIDIA Jetson - Video analytics on your existing CCTV and IP cameras: counting, zones, events, alerts - Quality control and defect detection on production lines - Camera and lens selection calculated for your distance and object size - Auto-labeling and retraining, so accuracy improves instead of drifting ⚙️ FIRMWARE, MOTION CONTROL & HARDWARE INTEGRATION - ESP32 firmware in Embedded C/C++: sensors, Wi-Fi, Bluetooth LE, OTA updates - Motion control: stepper motors, synchronized axes, camera-guided platforms - Integration of controllers, sensors, cameras, electronic locks, printers and scanners - Hardware with poor documentation or closed firmware - Software tested on a model of the device before the first boards arrive 📡 CONNECTED DEVICES, BLE APPS & FLEET MANAGEMENT - Flutter companion apps for iOS & Android with a stable BLE connection in the background - BLE protocol design, device registration and firmware updates through the app - Device-to-cloud integration: REST API, webhooks, MQTT - Fleet dashboard: status and version of every device, remote updates of firmware, models and services - Opening a closed device to partner apps 🔧 WHERE I USUALLY JOIN 💡 Idea or early prototype: feasibility, architecture and component choice before you buy hardware 🧩 Stalled project: the previous developer left; I audit the code, keep what works and finish it 🚀 Working prototype: from one unit on the bench to devices in the field, with remote updates and monitoring 📂 MY SELECTED PROJECTS - Network of camera stations: YOLO on Jetson at every station, automatic labeling and retraining, remote updates from one admin panel - Heavy rotating platform on stepper motors: moved from manual control to automatic camera tracking (ESP32, Jetson, industrial cameras) - Wearable device for a startup: Flutter app, BLE protocol, background reconnect and OTA firmware updates - Portable imaging device: DICOM import, AI image analysis, EMR integration and device management - Self-service luggage lockers and parcel lockers: lock control, touchscreen kiosk, delivery-partner integration 💻 TECH STACK 👁️ Vision & AI: Python, OpenCV, YOLO, PyTorch, TensorFlow, CUDA, NVIDIA Jetson, ONNX Runtime ⚙️ Embedded: Embedded C/C++, ESP32, stepper motor control, industrial cameras, sensors, ROS 2 📡 Connectivity: Bluetooth Low Energy, MQTT, REST, WebSockets, RTSP, OTA updates 📱 Apps: Flutter, Dart, Swift, Kotlin, .NET MAUI 🗄️ Backend & Cloud: C#, .NET, ASP.NET Core, FastAPI, PostgreSQL, Docker, Kubernetes, Azure, AWS 🏥 Medical: DICOM, EMR integration 📩 Send me what you're building: a short description, a schematic, a sample video or a repository. I'll reply with how I'd approach it, the main risks and what I'd do first.