SAN FRANCISCO · AI / ROBOTICS / SYSTEMS

Building AI systems for the physical world.

I’m Akshay Chalana, co-founder and CEO of Saphira. I work at the intersection of AI agents, engineering systems, and the infrastructure needed to reason about complex physical products.

01 / ABOUT

From reliable software to intelligent physical systems.

My work has moved through hardware reliability at Tesla, ML infrastructure at Meta and Apple, high-throughput trading systems at Citadel Securities, and now AI-native engineering at Saphira.

The common thread is building dependable systems in domains where complexity and correctness matter.

02 / SELECTED EXPERIENCE

A technical path through AI, infrastructure, and physical systems.

2021 — 2024NEW YORK

Citadel Securities Systems Software Engineer

Led roadmaps, mentored engineers, and owned high-reliability front-office trading systems across the software lifecycle, consolidating legacy applications and enabling roughly 1,000× growth in daily trade throughput.

2019 — 2021SEATTLE

Apple Software Engineer, ML Platform

Built distributed systems and low-latency analytics infrastructure for moving, processing, and understanding ML experiment data.

2018MENLO PARK

Meta PyTorch / ONNX

Designed and implemented the ONNX opset version converter for evolving model operators.

2016PALO ALTO

Tesla Reliability / Test Automation

Built software and firmware for hardware reliability testing and engineering test-data visualization.

03 / TECHNICAL WORK

Notes and working artifacts.

A home for technical work that is easier to understand through code, diagrams, and results than through résumé bullets.

CASE STUDY · FORTHCOMING

Engineering agents for complex systems

What we learned building agents that reason across requirements, architecture, failure analyses, tests, standards, and simulation outputs.

In preparation

EXPERIMENT · IN PROGRESS

Evaluation loops for physical AI

Exploring how system context and simulation access can help an agent identify what could break and construct meaningful evaluations.

Work in progress

WRITING · OPEN SLOT

Next technical note

This site is intentionally small. Substantive technical artifacts will replace this placeholder as they are ready.

Coming next

04 / CONTACT

Interested in the same hard problems?

I’m always glad to compare notes with people building in AI, robotics, simulation, and complex engineering systems.

akshay@chalana.com