About ayushv

I stripped down compute hardware to understand latency limits. From optimising RC telemetry to tuning inference kernels, the obsession remains: reducing overhead. Today, I apply that to HFT execution paths and agentic orchestration.

That early fascination with mechanics and optimization naturally evolved into a passion for software engineering, AI, and quantitative finance. I realized that code offers the ultimate playground for this statistical mindset: a place where I can dissect complex systems, optimize their performance, and build robust solutions that feel alive. Today, whether it's an AI agent or a high-frequency trading engine, I chase that same feeling of pure efficiency and control.

My path was a little unusual: AI came first, then I learned to code deeply. Today I still use models (Gemini/Claude/OpenAI) for drafting and validation, but I don’t ship blindly — I own the architecture and logic, and every critical path gets audits, targeted tests, and runtime checks before anything goes to production.

Beyond the code, I am deeply fascinated by the complex web of financial markets—the "cycles" as Ray Dalio describes them. This interest feels instinctive, rooted in my upbringing within a family jewellery business. I learned early on that the difference between profit and loss often wasn't in the craftsmanship, but in timing the market—knowing exactly when to stock up on gold and silver, and when to hedge. Today, I apply that same pragmatic rigor to building systems that don't just predict price, but capture value.

Getting the GitHub Student Developer Pack was a catalyst—access to professional tools like DigitalOcean and JetBrains allowed me to move from localhost experiments to deploying live production systems.