About Amandeep
TL;DR — I'm Amandeep - a systems-oriented software engineer. I enjoy working at lower levels of software - parallel computation, memory behavior, GPU pipelines, and computational geometry — and turning that understanding into reliable, production-grade software. I care deeply about how systems behave under real constraints, and how good engineering choices compound at the product level.
Engineer, writer, and meditator - building real-time systems and exploring how we think, decide, and create.
I'm Amandeep Singh — a software engineer drawn to problems where systems, math, and real-world constraints intersect. Over the last decade, my work has revolved around building performance‑sensitive systems: understanding how computation flows through CPUs and GPUs, how memory layout and parallelism shape outcomes, and how small architectural decisions ripple through an entire product.
Much of my experience comes from working in constrained environments — browsers, mobile devices, inconsistent sensors, and unpredictable hardware. These limits force clarity. They push you to reason about SIMD, threading models, cache behavior, GPU pipelines, and numerical trade‑offs — not as theory, but as tools for building systems that actually hold up in production.
What excites me most is the layer beneath features: computational geometry, computer vision, rendering and compute pipelines, and the invisible infrastructure that makes higher‑level experiences feel stable and trustworthy. I enjoy bridging R&D with product — taking ideas from first principles to shipping systems that teams and users can rely on.
Outside of engineering, I spend time thinking about how people think. As a long‑time Heartfulness meditation practitioner and trainer, I care about clarity — in attention, decision‑making, and design. Writing and meditation are how I slow things down, reflect, and bring the same intentionality into both technical and human systems.
I have been fortunate to have my work recognized by senior R&D leadership, including Dr. Mark Hodgetts (Oxford & Cambridge PhD), for original contributions to performance‑critical systems spanning geometry, GPU optimization, and real‑time pipelines.
What I'm Working On
- Systems‑level engineering — parallel computation, memory models, SIMD, and designing software that behaves predictably under load and constraint.
- GPU & compute pipelines — reasoning about data‑parallel workloads, shader‑friendly algorithms, and performance trade‑offs across CPU and GPU boundaries.
- Computational geometry & vision — the math and representations that underpin robust spatial and visual systems.
- Product‑minded architecture — turning low‑level capabilities into stable APIs, primitives, and systems that scale across teams and use cases.
- Thinking clearly — writing, mentoring, and Heartfulness meditation as practices for better decisions, better systems, and better leadership.
If something here resonates — whether you are building a product, learning, hiring, or just curious — feel free to reach out. I am always happy to talk about systems, AR/XR, computational graphics, or the discipline of thinking clearly.