Senior Software Architect & Engineer

Company Description KAPDAA is a London-based sustainable recycling company that partners with fashion brands and councils to transform UK garment waste into valuable recycled products.. We are currently developing AI4Fibres, an innovative AI-powered textile recycling system.Role Overview We are hiring a Senior Software Architect & Engineer to architect, build, and operate a real-time, high-throughput computer vision system at production scale and the software platform around it. The core challenge is system design for real-time multi-camera streams and processing: an optimized C++ vision pipeline handling a continuous stream of items under strict latency budgets, and the architecture built around it the right choice of protocols, optimisations, and designs to handle high throughput in real time, from capture at the edge through processing, storage, and sync to the cloud. You will own the architecture end to end and justify design choices in terms of throughput, latency, cost, and failure modes.Role DescriptionSystem architecture: design the production-scale, real-time vision platform edge devices, cameras, GPU inference, storage, edge→cloud pipeline around throughput and latency constraints.Real-time C++ core: multi-camera capture and processing, GPU inference integration, concurrency and memory design that holds the per-garment latency budget.APIs & data flow: gRPC/Protocol Buffers, streaming, message queues; batching and backpressure sized to real production volumes.Storage & pipelines: high-volume image storage (retention, tiering, cost control),PostgreSQL for high-ingest operational data, resilient edge→cloud sync (Kafka + Debezium CDC, Redis).UI & web layer: React/Next.js operator dashboards with live data views; FastAPI services (REST, validation, authentication, WebSockets).Deployment & operations: Docker, Ansible, GitHub Actions, Azure (ACR); versioned rollouts, monitoring, and remote diagnostics across the edge fleet; debug latency and throughput issues end to end.QualificationsSystem design distributed, real-time architecture proven in production: capacity planning, latency budgeting, graceful degradation, failure handling.Modern C++ (17+) & performance engineering profiling, multithreading/concurrency, CPU/memory awareness.Linux & edge strong fundamentals, Bash, troubleshooting, production operations on GPU-backed edge devices (e.g., NVIDIA Jetson).gRPC + Protocol Buffers, REST, WebSockets; Kafka + Debezium (CDC), RedisVision/video pipelines: OpenCV, GStreamer/DeepStream; GPU inference integration (CUDA/TensorRT) depth a strong plusPostgreSQL (schema design, indexing, optimization), SQLAlchemy; object storage for image dataReact, Next.js, TypeScript, Tailwind; live/streaming data UIsPython, FastAPI, Pydantic, authentication, security basicsDocker, GitHub Actions, Ansible, Azure (ACR + container deployments)AI-assisted development: skilled with coding agents and AI tools (Claude Code, Codex, LLM-based workflows) in everyday engineer.We're looking for an engineer who owns the architecture and still writes the code, justify those decisions in terms of throughput, latency, cost and failure modes, and then prove them in the hot path under load. You'll hold a high engineering bar (code reviews, secure coding, meaningful load and latency tests, reliable CI/CD) and use AI coding tools without lowering it. Just as importantly, you'll explain your design clearly to both the implementation team and to management, and you'll be comfortable in a fast-moving start-up where requirements shift, ownership is broad, and the system you build has to keep running live.