Shivansh S.

I build systems that hold up in production - backend services, AI agents, and robotics software. Not demos. What I do: BACKEND & INFRASTRUCTURE Services and APIs in Go, Python, and Node/TypeScript. PostgreSQL, REST and gRPC. Docker, AWS, CI/CD, and real observability with Prometheus and Grafana. AI AGENTS & AUTOMATION Agents and RAG pipelines with LangChain, OpenAI and Claude APIs, and MCP servers - plus the evaluation to prove retrieval actually works. Workflow automation with n8n and custom integrations. PERFORMANCE-CRITICAL & EMBEDDED C++ Real-time data paths, lock-free concurrency, and profiling. ROS2, computer vision, and CUDA-accelerated inference when the problem calls for it. BACKGROUND BASc in Electrical & Computer Engineering, University of Toronto, with minors in AI Engineering and Robotics & Mechatronics. A year of production C++ on autonomous cargo aircraft. Software lead on an autonomous multicopter. Perception acceleration on a full-scale self-driving car. I've published an extension to the Visual Studio Marketplace and I run a multi-agent system that manages my own workflows daily. The common thread: most projects fail the same way - nothing was instrumented, so when it breaks nobody can say why. I build the logging, tracing and tests in from the start. I would rather scope something small and deliver early than promise a platform. Tell me what breaks today and I'll tell you honestly whether I'm the right fit.